Demo20200822.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "name": "Demo20200822.ipynb",
      "provenance": [],
      "collapsed_sections": []
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    },
    "accelerator": "GPU"
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "upc8E6xK49py",
        "colab_type": "text"
      },
      "source": [
        "## 信用卡詐欺"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "jw4UQGIm44B8",
        "colab_type": "text"
      },
      "source": [
        "### 資料下載"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "RprQZFgP4IDe",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 228
        },
        "outputId": "4106246d-6b57-4980-d927-a1d4ca275719"
      },
      "source": [
        "! wget https://raw.githubusercontent.com/ywchiu/tibamedl/master/Data/creditcardfraud.zip"
      ],
      "execution_count": 1,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "--2020-08-22 01:58:24--  https://raw.githubusercontent.com/ywchiu/tibamedl/master/Data/creditcardfraud.zip\n",
            "Resolving raw.githubusercontent.com (raw.githubusercontent.com)... 151.101.0.133, 151.101.64.133, 151.101.128.133, ...\n",
            "Connecting to raw.githubusercontent.com (raw.githubusercontent.com)|151.101.0.133|:443... connected.\n",
            "HTTP request sent, awaiting response... 200 OK\n",
            "Length: 71387154 (68M) [application/zip]\n",
            "Saving to: ‘creditcardfraud.zip’\n",
            "\n",
            "creditcardfraud.zip 100%[===================>]  68.08M   270MB/s    in 0.3s    \n",
            "\n",
            "2020-08-22 01:58:27 (270 MB/s) - ‘creditcardfraud.zip’ saved [71387154/71387154]\n",
            "\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "4kq7Mo2c40s-",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "! unzip creditcardfraud.zip > 1.log"
      ],
      "execution_count": 2,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "u-zJq6uJ5EXa",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 35
        },
        "outputId": "5416418c-564e-418f-d707-23fba5b34302"
      },
      "source": [
        "! ls"
      ],
      "execution_count": 3,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "1.log  creditcard.csv  creditcardfraud.zip  sample_data\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "JagV7KoK5KHf",
        "colab_type": "text"
      },
      "source": [
        "### 讀取資料"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "ToojJiWd5Jen",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "import pandas\n",
        "df = pandas.read_csv('creditcard.csv')"
      ],
      "execution_count": 4,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "OjJRlTQp5GPJ",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 222
        },
        "outputId": "6fa74fb9-42b9-4ede-cf3c-ae4960306612"
      },
      "source": [
        "df.head()"
      ],
      "execution_count": 5,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/html": [
              "<div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>Time</th>\n",
              "      <th>V1</th>\n",
              "      <th>V2</th>\n",
              "      <th>V3</th>\n",
              "      <th>V4</th>\n",
              "      <th>V5</th>\n",
              "      <th>V6</th>\n",
              "      <th>V7</th>\n",
              "      <th>V8</th>\n",
              "      <th>V9</th>\n",
              "      <th>V10</th>\n",
              "      <th>V11</th>\n",
              "      <th>V12</th>\n",
              "      <th>V13</th>\n",
              "      <th>V14</th>\n",
              "      <th>V15</th>\n",
              "      <th>V16</th>\n",
              "      <th>V17</th>\n",
              "      <th>V18</th>\n",
              "      <th>V19</th>\n",
              "      <th>V20</th>\n",
              "      <th>V21</th>\n",
              "      <th>V22</th>\n",
              "      <th>V23</th>\n",
              "      <th>V24</th>\n",
              "      <th>V25</th>\n",
              "      <th>V26</th>\n",
              "      <th>V27</th>\n",
              "      <th>V28</th>\n",
              "      <th>Amount</th>\n",
              "      <th>Class</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>0.0</td>\n",
              "      <td>-1.359807</td>\n",
              "      <td>-0.072781</td>\n",
              "      <td>2.536347</td>\n",
              "      <td>1.378155</td>\n",
              "      <td>-0.338321</td>\n",
              "      <td>0.462388</td>\n",
              "      <td>0.239599</td>\n",
              "      <td>0.098698</td>\n",
              "      <td>0.363787</td>\n",
              "      <td>0.090794</td>\n",
              "      <td>-0.551600</td>\n",
              "      <td>-0.617801</td>\n",
              "      <td>-0.991390</td>\n",
              "      <td>-0.311169</td>\n",
              "      <td>1.468177</td>\n",
              "      <td>-0.470401</td>\n",
              "      <td>0.207971</td>\n",
              "      <td>0.025791</td>\n",
              "      <td>0.403993</td>\n",
              "      <td>0.251412</td>\n",
              "      <td>-0.018307</td>\n",
              "      <td>0.277838</td>\n",
              "      <td>-0.110474</td>\n",
              "      <td>0.066928</td>\n",
              "      <td>0.128539</td>\n",
              "      <td>-0.189115</td>\n",
              "      <td>0.133558</td>\n",
              "      <td>-0.021053</td>\n",
              "      <td>149.62</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>0.0</td>\n",
              "      <td>1.191857</td>\n",
              "      <td>0.266151</td>\n",
              "      <td>0.166480</td>\n",
              "      <td>0.448154</td>\n",
              "      <td>0.060018</td>\n",
              "      <td>-0.082361</td>\n",
              "      <td>-0.078803</td>\n",
              "      <td>0.085102</td>\n",
              "      <td>-0.255425</td>\n",
              "      <td>-0.166974</td>\n",
              "      <td>1.612727</td>\n",
              "      <td>1.065235</td>\n",
              "      <td>0.489095</td>\n",
              "      <td>-0.143772</td>\n",
              "      <td>0.635558</td>\n",
              "      <td>0.463917</td>\n",
              "      <td>-0.114805</td>\n",
              "      <td>-0.183361</td>\n",
              "      <td>-0.145783</td>\n",
              "      <td>-0.069083</td>\n",
              "      <td>-0.225775</td>\n",
              "      <td>-0.638672</td>\n",
              "      <td>0.101288</td>\n",
              "      <td>-0.339846</td>\n",
              "      <td>0.167170</td>\n",
              "      <td>0.125895</td>\n",
              "      <td>-0.008983</td>\n",
              "      <td>0.014724</td>\n",
              "      <td>2.69</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>1.0</td>\n",
              "      <td>-1.358354</td>\n",
              "      <td>-1.340163</td>\n",
              "      <td>1.773209</td>\n",
              "      <td>0.379780</td>\n",
              "      <td>-0.503198</td>\n",
              "      <td>1.800499</td>\n",
              "      <td>0.791461</td>\n",
              "      <td>0.247676</td>\n",
              "      <td>-1.514654</td>\n",
              "      <td>0.207643</td>\n",
              "      <td>0.624501</td>\n",
              "      <td>0.066084</td>\n",
              "      <td>0.717293</td>\n",
              "      <td>-0.165946</td>\n",
              "      <td>2.345865</td>\n",
              "      <td>-2.890083</td>\n",
              "      <td>1.109969</td>\n",
              "      <td>-0.121359</td>\n",
              "      <td>-2.261857</td>\n",
              "      <td>0.524980</td>\n",
              "      <td>0.247998</td>\n",
              "      <td>0.771679</td>\n",
              "      <td>0.909412</td>\n",
              "      <td>-0.689281</td>\n",
              "      <td>-0.327642</td>\n",
              "      <td>-0.139097</td>\n",
              "      <td>-0.055353</td>\n",
              "      <td>-0.059752</td>\n",
              "      <td>378.66</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>1.0</td>\n",
              "      <td>-0.966272</td>\n",
              "      <td>-0.185226</td>\n",
              "      <td>1.792993</td>\n",
              "      <td>-0.863291</td>\n",
              "      <td>-0.010309</td>\n",
              "      <td>1.247203</td>\n",
              "      <td>0.237609</td>\n",
              "      <td>0.377436</td>\n",
              "      <td>-1.387024</td>\n",
              "      <td>-0.054952</td>\n",
              "      <td>-0.226487</td>\n",
              "      <td>0.178228</td>\n",
              "      <td>0.507757</td>\n",
              "      <td>-0.287924</td>\n",
              "      <td>-0.631418</td>\n",
              "      <td>-1.059647</td>\n",
              "      <td>-0.684093</td>\n",
              "      <td>1.965775</td>\n",
              "      <td>-1.232622</td>\n",
              "      <td>-0.208038</td>\n",
              "      <td>-0.108300</td>\n",
              "      <td>0.005274</td>\n",
              "      <td>-0.190321</td>\n",
              "      <td>-1.175575</td>\n",
              "      <td>0.647376</td>\n",
              "      <td>-0.221929</td>\n",
              "      <td>0.062723</td>\n",
              "      <td>0.061458</td>\n",
              "      <td>123.50</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>2.0</td>\n",
              "      <td>-1.158233</td>\n",
              "      <td>0.877737</td>\n",
              "      <td>1.548718</td>\n",
              "      <td>0.403034</td>\n",
              "      <td>-0.407193</td>\n",
              "      <td>0.095921</td>\n",
              "      <td>0.592941</td>\n",
              "      <td>-0.270533</td>\n",
              "      <td>0.817739</td>\n",
              "      <td>0.753074</td>\n",
              "      <td>-0.822843</td>\n",
              "      <td>0.538196</td>\n",
              "      <td>1.345852</td>\n",
              "      <td>-1.119670</td>\n",
              "      <td>0.175121</td>\n",
              "      <td>-0.451449</td>\n",
              "      <td>-0.237033</td>\n",
              "      <td>-0.038195</td>\n",
              "      <td>0.803487</td>\n",
              "      <td>0.408542</td>\n",
              "      <td>-0.009431</td>\n",
              "      <td>0.798278</td>\n",
              "      <td>-0.137458</td>\n",
              "      <td>0.141267</td>\n",
              "      <td>-0.206010</td>\n",
              "      <td>0.502292</td>\n",
              "      <td>0.219422</td>\n",
              "      <td>0.215153</td>\n",
              "      <td>69.99</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "   Time        V1        V2        V3  ...       V27       V28  Amount  Class\n",
              "0   0.0 -1.359807 -0.072781  2.536347  ...  0.133558 -0.021053  149.62      0\n",
              "1   0.0  1.191857  0.266151  0.166480  ... -0.008983  0.014724    2.69      0\n",
              "2   1.0 -1.358354 -1.340163  1.773209  ... -0.055353 -0.059752  378.66      0\n",
              "3   1.0 -0.966272 -0.185226  1.792993  ...  0.062723  0.061458  123.50      0\n",
              "4   2.0 -1.158233  0.877737  1.548718  ...  0.219422  0.215153   69.99      0\n",
              "\n",
              "[5 rows x 31 columns]"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 5
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "9D8d3btc5qCR",
        "colab_type": "text"
      },
      "source": [
        "### EDA (Exploratory Data Analysis)"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "GCJ0Qci-5QHX",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "count_classes = pandas.value_counts(df['Class'], sort = True)"
      ],
      "execution_count": 7,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "96qw8zpl5uoA",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 69
        },
        "outputId": "3d7609fb-63bf-4355-a31c-67905de75e7b"
      },
      "source": [
        "count_classes"
      ],
      "execution_count": 8,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "0    284315\n",
              "1       492\n",
              "Name: Class, dtype: int64"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 8
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "jSVfSDFg5xYB",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 283
        },
        "outputId": "1ea514c3-e017-4b4d-cd84-5a87d40bbd18"
      },
      "source": [
        "count_classes.plot(kind = 'bar', rot=0)"
      ],
      "execution_count": 9,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<matplotlib.axes._subplots.AxesSubplot at 0x7f0217376828>"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 9
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "EJ4avNFQ54ju",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "frauds = df[df.Class == 1]\n",
        "normal = df[df.Class == 0]"
      ],
      "execution_count": 10,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "aCoepQ186HWs",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 308
        },
        "outputId": "d048a23c-f599-48b5-e13a-c027c2b96851"
      },
      "source": [
        "from matplotlib import pyplot as plt\n",
        "f, (ax1, ax2) = plt.subplots(2, 1, sharex=True)\n",
        "f.suptitle('Amount per transaction by class')\n",
        "\n",
        "bins = 50\n",
        "\n",
        "ax1.hist(frauds.Amount, bins = bins)\n",
        "ax1.set_title('Fraud')\n",
        "\n",
        "ax2.hist(normal.Amount, bins = bins)\n",
        "ax2.set_title('Normal')\n",
        "\n",
        "plt.xlabel('Amount ($)')\n",
        "plt.ylabel('Number of Transactions')\n",
        "plt.xlim((0, 20000))\n",
        "plt.yscale('log')\n",
        "plt.show()"
      ],
      "execution_count": 12,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 2 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Gh0gF9Ib6OuL",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 308
        },
        "outputId": "45b1567c-d109-470a-94f7-2cd96f53f487"
      },
      "source": [
        "f, (ax1, ax2) = plt.subplots(2, 1, sharex=True)\n",
        "f.suptitle('Time of transaction vs Amount by class')\n",
        "\n",
        "ax1.scatter(frauds.Time, frauds.Amount)\n",
        "ax1.set_title('Fraud')\n",
        "\n",
        "ax2.scatter(normal.Time, normal.Amount)\n",
        "ax2.set_title('Normal')\n",
        "\n",
        "plt.xlabel('Time (in Seconds)')\n",
        "plt.ylabel('Amount')\n",
        "plt.show()"
      ],
      "execution_count": 13,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAZkAAAEjCAYAAADg9HzdAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4yLjIsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+WH4yJAAAgAElEQVR4nO29e5gdVZX//fl2pwlNuHQikSENIYgYXjRKJCNxoo6oQxBFMnhBRoRxfGXmp6OCmjH85BFEfMmYUZG56OjICIIYEGxR1MgI6MgYJCGBGCFDUBJobpHQ3NJCp7PeP2pXp/p01Tl1Tp/qc06f9XmefrrOrtvaVbv22nvttdeWmeE4juM4RdDRaAEcx3GcyYsrGcdxHKcwXMk4juM4heFKxnEcxykMVzKO4zhOYbiScRzHcQrDlUwLI2mjpNc3gRz/R9Kjkp6R9IJGy1MEkmaH/HU2WpbJjqTzJV1R8D3+WtIvi7yHE+FKpokJlVr8t0vSYOL3e8zspWZ2S4Nl7AK+CBxnZnub2eMl++dIMklTGiNhbUi6X9Kb4t9mtjXkb7iRclUiVNAm6ZhGy5JGq5YHp3ZcyTQxoVLb28z2BrYCJybSrmy0fIEDgD2BjbVewCuc+iBJwOnA9vDfcRqOK5kWJtnaDi3YayRdIelpSRskvUTSOZIek/SApOMS5+4n6RuSHpbUL+nCLFOQpKmSLpb0UPi7OKS9BNgUDhuQdFPK6b9I7H9G0quDqeJWSV+S9DhwvqTDJN0k6XFJf5B0paSekrx+QtJdkp6UtFLSnmHf/pJ+KGlA0nZJ/y2pI+xbJum+8Ex+K+kvS/L2AUl3J/a/UtK3gNnAD4LM/1DaApc0S9L14X6bJX0gcc3zJV0t6fJw3Y2SFmQ8269I+qeStO9L+ljY/mR4P09L2iTpjWnXCbwWOBD4CPBuSXskrpl85gOSfifpz0L6A6GMnJE4fr8g/zZJWySdm3imo8xZKc/mFkmfDfd7WtJPJe2fVR4y8rJneMdPS7pD0ivCtZdKurbkeV0i6csZz/dgSdeFfDwu6V8yjvtyeA5PSVor6bWJfa+StCbse1TSF0P6noq+t8fDM71d0gEZ+WlfzMz/WuAPuB94U1YacD7wR2AxMAW4HPg98CmgC/gA8PvEud8D/h2YBrwQ+DXwtxn3vgBYHY6bCfwP8Nmwbw5gwJSMc8fsB/4a2Al8OMjaDbwY+AtgarjHL4CLS/L6a2AWMAO4G/i7sO8i4Kshn11Ela3CvneGczqAU4BngQMT+/qBPwUUZDgk7XmX5iPI929EvbijgG3AG0rexQlAZ5BvdcbzeR3wQELe6cBgkHlu2DcrIcNhZcrIN4CrwzN4HHh7yjN/X5DpQqLe8b+GZ34c8DSwdzj+cuD7wD7hvv8LvD+RvyvKPJtbgPuAl4R3ewuwPE95SVx/CHhHyMsniMpyF5ESfRboCcdOAR4Djk65TidwJ/AlonK+J/CaxPP4ZeLY04AXhOt9HHgE2DPs+xXw3rC9N7AwbP8t8ANgr3Cvo4F9G11XNNtfwwXwv5wvKp+SuTGx70TgGaAz/N4nfNw9RCau54DuxPGnAjdn3Ps+4ITE78XA/WG7bKWRtj984Fsr5HcJsK4kr6clfn8e+GrYvoCoQnxxjue4HjgpbK8CPprneSfzARwMDAP7JPZfBHwz8S7+K7HvSGAw4z4iquxfF35/ALgpbL+YqAJ9E9BVIV97AU8BS8Lvfwe+X/LM7038nhfyc0Ai7XEihdkJPA8cmdj3t8AtifxVUjLnJvZ/EPhJnvKSuP7qxO8O4GHgteH3j4EPhO23Ar/NuM6riZT/mHtRomRS9j8BvCJs/wL4DLB/yTF/Q9Tgenk133K7/bm5bHLxaGJ7EPiD7R6oHgz/9wYOIWoVPhy6+QNEldILM647C9iS+L0lpI2HB5I/JB0g6TvBNPQUcAWwf8k5jyS2dxDlBWAFsBn4aTADLUtc93RJ6xP5fFniugcTKdBqmQVsN7OnE2lbgN4ysu6plLEni2qr7xApeYC/Aq4M+zYDZxFVuo+F55P13P+SqKfyo/D7SuDNkmYmjiktH5hZadreRM+ni7HvPJm/SmS9q7yMlA8z2wU8yO4ydxlRz4Pw/1sZ1zgY2GJmOyvdLJhi7w6m2AFgP3aXk/cT9cruCSaxt4b0bxE1VL6jyIz8eUWOME4CVzLtyQNEPZn9zawn/O1rZi/NOP4hIsUUMzuk5SErzHdp+v8X0uaZ2b5ElYdy3cDsaTP7uJm9CHgb8DFJb5R0CPB14O+BF5hZD/CbxHUfAA6rUm6I8j5D0j6JtNlEprdauAp4R5D3GGBkzMHMvm1mryF6/gb8Y8Y1ziCqyLdKegS4hkhR/FUN8vyByFxV+s7j/D1L1HOK+ZMqrp037PvB8UYYCzqI3WWuD3i5pJcR9WSynGAeAGanKfckYfzlH4B3AdNDOXmSUE7M7F4zO5WoEfaPwHclTTOzITP7jJkdCfxZkMUdLkpwJdOGmNnDwE+BL0jaV1KHooH3P8845SrgXEkzwwDup4l6GnnYBuwCXlThuH2IzHtPSuoFlua8PpLeKunFkkRUOQyHe04jqtS2hePeR9STifkP4BOSjlbEi0NFD1GrP1VmM3uAyExyURj8fTlRa7emuR1mto6oYv8PYJWZDQR550p6g6SpRGM8gyFfpfnvBd5IVMkdFf5eQVQhVl3phd7v1cDnJO0TnsnH2J2/9cDrFM0d2g84p4rL5y0PR0s6OSiIs4gaRauDfH8Evgt8G/i1mW3NuMavicxsyyVNC+9qUcpx+xD1ArcBUyR9Gtg33inpNEkzQ49qICTvknSspHmKHGaeIlLMY95Pu+NKpn05HdgD+C2R/fm7RIOqaVwIrAHuAjYAd4S0ipjZDuBzwK3BZLUw49DPAK8kUhI3ANflywYAhwP/RaSkfgX8m5ndbGa/Bb4Q0h4lGoe4NSHbNUG2bxMNevcRORVANMZybpD5Eyn3PJVofOEhIieK88zsv6qQuZRvE429fDuRNhVYTqSAHiFqSadV6O8F1pvZT83skfgPuITdLf5q+TBRj+V3wC+DXJcCmNmNwEqi8rAW+GHei1ZRHr5P5KjxRMjfyWY2lNh/GdH7zDKVxcryRKKxra1EJrdTUg5dBfyEyLlhC5FCT5pzjwc2SnoG+DLwbjMbJOrBfZdIwdwN/LycPO1K7NHiOI7TMkiaDdwD/ImZPdVoeZxsvCfjOE5LEcZoPgZ8xxVM8+MzrR3HaRkkTSMyfW4hMmM5TY6byxynhZH0TeBBMzu30bI4ThpuLnOcOqEo9E0yiOkzZea1OE5b4ErGcepLMojp3mY2Mp+o0nwNx5mMuJJxnAJRFDjyQ5LuBe4NaeWCMX5T0oWJ36+X9GDi93xFASOflrSSKB6X4zQtrmQcp3iWEM3kPzL8vp1owuQMovkn1yhElC6HoqjKfURzMWYQzep/exECO069cCXjOPWlL46TJqkvpF1kZtvDBD7M7Aoze9zMdprZF4gmXc7Nce2FRKFiLg4hTb5LpLAcp2lxG7Hj1JclyZn/koyxwUA/QRSGZhZR2Jt9GRsMNI1ZQL+NdgndknWw4zQD3pNxnOIZUQqVgjFSPvjkw0BviNEWM7sQiR2nTriScZyJpWwwRqLgkydImiHpT4iCQ8b8Kpz7EUldkk4GXjVBcjtOTbiScZyJpVIwxm8RreZ4P1Gk7JXxDjN7HjiZaMGt7UTBHqsJJOo4E47P+Hccx3EKw3syjuM4TmG4knEcx3EKw5WM4ziOUxiuZBzHcZzCaLvJmPvvv7/NmTOn0WI4juO0DGvXrv2Dmc2s5dy2UzJz5sxhzZo1jRbDcRynZZBUc2SJtlMyjuM0F33r+lmxahMPDQwyq6ebpYvnsmR+b6PFcuqEKxnHcRpG37p+zrluA4NDwwD0DwxyznUbAFzRTBJ84N9xnIaxYtWmEQUTMzg0zIpVmxokkVNvXMk4jtMwHhoYrCrdaT1cyTiO0zBm9XRXle60HhOuZCQdLOlmSb+VtFHSR0P6DEk3Sro3/J8e0iXpEkmbJd0l6ZWJa50Rjr9X0hkTnRfHccbH0sVz6e7qHJXW3dXJ0sV51nBzWoFG9GR2Ah83syOJVvr7kKQjgWXAz8zscOBn4TfAm4HDw9+ZwFcgUkrAeUTL2r4KOC9WTI7jtAZL5vdy0cnz6O3pRkBvTzcXnTzPB/0nERPuXWZmDxMtvoSZPS3pbqAXOAl4fTjsMuAW4JMh/fKwGuBqST2SDgzH3mhm2wEk3QgcD1w1YZlxHGfcLJnf60plEtPQMRlJc4D5wG3AAUEBATwCHBC2exm93saDIS0rPe0+Z0paI2nNtm3b6ia/4ziOU56GKRlJewPXAmeZ2VPJfaHXUreFbszsa2a2wMwWzJxZU2QEx3EcpwYaomQkdREpmCvNLF7Z79FgBiP8fyyk9wMHJ04/KKRlpTuO4zhNQiO8ywR8A7jbzL6Y2HU9EHuInQF8P5F+evAyWwg8Gcxqq4DjJE0PA/7HhTTHcRynSWhEWJlFwHuBDZLWh7T/CywHrpb0fqK1z98V9v0IOAHYDOwA3gdgZtslfRa4PRx3QewE4DiO4zQHioY/2ocFCxbYREdh9gCAjuO0MpLWmtmCWs71AJkF4wEAHcdpZzysTMF4AEDHcdoZVzIF4wEAHcdpZ1zJFIwHAHQcp51xJVMwHgDQcZx2xgf+CyYe3HfvMsdx2hFXMhOABwB0HKddcXOZ4ziOUxiuZBzHcZzCcCXjOI7jFIYrGcdxHKcwXMk4juM4heHeZY7jOAXS7gFyXck4juMUhAfIdXOZ4zhOYXiAXFcyjuM4heEBct1c5kxi2t0W7jSeWT3d9KcolHYKkOtKxml50pQJ0Pa2cKfxLF08d1Q5hPYLkOvLLzstTenAKkQf8Z5dHTyxY2jM8T3dXUybOsV7N86EMRl61L78stO2ZA2slqbFDAwOMTAYKR/v3UwumrUyb/cAuT7w77Q04x1AbTdPn8lK3KPtHxjE2N2A6FvX32jR2h5XMk5LkzWA2tPdNWaxuCzaydNnsuKuws2LKxmnpclaefT8t72Ui06eR29PNwJ6e7qZvldX6jXaydNnsuKuws2Lj8k4LU2llUeTtvAsJ4F28vSZrLircPPiSqYFaNYBzWYh78CqL4U9eXFX4ebFlUyT47GP6ku7e/pMVrwB0by4kmlyyg1o+gfkOLvxBkRz4gP/TY4PaDqO08q4kmlysgYufUDTcZxWwJVMk5PlousDmo7TGvSt62fR8ps4dNkNLFp+U9tNEPUxmSbHBzQdp3Vxxx1XMi2BD2g6TmvijjtuLnMcxykMd9xpkJKRdKmkxyT9JpE2Q9KNku4N/6eHdEm6RNJmSXdJemXinDPC8fdKOqMReXEcx8nCHXca15P5JnB8Sdoy4Gdmdjjws/Ab4M3A4eHvTOArECkl4DzgGOBVwHmxYnIcxymKagby3XGnQUrGzH4BbC9JPgm4LGxfBixJpF9uEauBHkkHAouBG81su5k9AdzIWMXlOI5TN6pdUmDJ/N4xgVovOnle24zHQHMN/B9gZg+H7UeAA8J2L/BA4rgHQ1pW+hgknUnUC2L27Nl1FNlxnHailoH8dnfcaSYlM4KZmaS6rQttZl8DvgbR8sv1uq7jOM1JUUFlfSC/eprJu+zRYAYj/H8spPcDByeOOyikZaU7jtPGFLlKpg/kV08zKZnrgdhD7Azg+4n004OX2ULgyWBWWwUcJ2l6GPA/LqQ5jtPGFLlKpg/kV09DzGWSrgJeD+wv6UEiL7HlwNWS3g9sAd4VDv8RcAKwGdgBvA/AzLZL+ixwezjuAjMrdSZwHKfNKNKk5RE4qqchSsbMTs3Y9caUYw34UMZ1LgUuraNojtPQReJ8gbrxU/Qqme0+kF8tzWQuc1qYyRIEsEh7fjPfezLhJq3moim9y5zdNHPLNpatf2AQAbHbXisHAWxkrCmPc1Uf3KTVXLiSaWKaOYJrqWylfuGtWjk20kXV3WPrh5u0mgdXMk1MvVu25XpF1faY0mQrpdkqxzx5LNqeX45G3rtVKH2Hxx4xk5vv2VZ1j6VvXT/nX7+RgcEhAKbv1cV5J77UFVMBuJJpYurZsi3XKwKq7jHlkaGZKse8vcKli+eOOg4mzp7fyHu3Amnv8IrVW0f25+3p963rZ+k1dzK0a3f/+4kdQyz97p0Vz3Wqxwf+m5h6Tvwq1yuqZV5BJRmarXLMm8dGxpryOFflydN7zjMfZsWqTaMUTMzQsI05d7I4tDQS78k0MfVs2dbSKyq3L022ePC/p7sLCc5euZ4VqzY1xaBrNflvpD3fxxKyyduDr3Rc3jLfzGOirYQrmSamnl4ylez91Y4FZMkG1ZveJgIf76gPjfR2zHqHacfVep3kuc3u7dfMnqdJXMk0OfVq2VbqFdXSY0qTbdHym5ryw5wM4x2NrlQa3bJPe4el5HmnSxfPHTMmA9DVqVHnNrO3X6PfRTX4mEybUM7eX8+xgGb9MFt9vKMZJmoWGRMsD2nv8LSFs6t+p0vm97Lina+gp7trVPreU0e3uZs5GGaj30U1eE+mjSjXK6pXj6mZzVJpeWx07yAvWZXK+ddvnDD5m6EBUa9yGl8j2Rt4YsfQqN5AM/d+m+Fd5MV7MjlxL5N8tFJIj2boHeQlq/IYGByaMPmbuWVfC5V6A83c+22ld+E9mRy0kv2z0dTLWWEiehiNGtg9t28DV932AMNmdEqceszBXLhkXtlz8g56Fyl/M7fsayFPb6BZvf1a6V24kslBs3uZNBvj/TAnSqlnVdpFmhzO7dswagLhsNnI73KKJs+gd0xR8jdrTLBaGyTNbNqtRLO+izRcyeSgleyfjaDevY5alXo1cvSt6x8V1DNJkZXMVbc9kJleTsmkVSo7nt/JEzuGxhxbpPzN1rIfT4OklXoDaTTbu8jClUwOWrnFUzRF9DoqKfU0ZQLVzc9ZsWpTqoIRFFrJDFvaXaP0RctvKqsgSyuV0mcPrVVJ1oPxWBlaqTfQyriSyUGrt3iKpAhTYjmlnqXU9uzqqEqOLEVmRJVPUWNCnVKmoonznFdReyU5fitDq/QGWhlXMjnwjzmbIkyJ5ZR6llLLGqvIkiNLkfWWUWQw/jGhU485eNSYTBbVtMbbuRy6lWE3zeqO70omJ+3+MWeR9ZHvVzLRrRxpH8dFJ89L/WDOXrm+avnSqEWR1cPRIx53SXqXZfVsfMwvnWR52a+7i65OMTS8+xm2o5WhmT1gXck442Lp4rl87Or1lAa1ffb5nfSt669YwLM+jotOnsety94w5vgspdbT3cVzO3flNmmW651mKbJ6VfoXLpnHgkNmjNw7S9EU2Rpv1lZvJUrLy8DgEF0dYvpeXQzsGBqVl0bmcaLunVydtpRm8YB1JeOMm5So6SNh0ysV8Gp7DVk9kLe+4kB+eOfDI+l5FqEqVTTxJLyevboK9doqrSjTFEyl1ngtc22y7t9Mrd5KpJWXoV3GXntMYd2njxtJqyWP9VrIbKKeb5rjRynN0Bt2JeOMi3KxkvIU8GrGdOJW2+DQ8Ejrvzesjnjt2v5RH9sfh3aNOi+t8oCxHmlLr7kztdKPgyfWo4VaaV0UAW8/Ots8W+tcm3L3b5ZWbyXylpdq81jPhcyqufd4ylOe9XWaYWzKlYwzLsopkjwFPO/AbVrrv6tD7Hh+Z+pAejI8SFblsffUKamt4jSm7RF9KvVooVZSvgbcfM+2zP21zrWpdP9maPVWIm95qTaPlRYyq8f7LU0fb4+n0vtqlrEpj13mjItyiuTYI2ZWPL9SrLM4ZtxZK9enKoQ0s1ZMbALLqjzKnVvKk4NDmS3Uj199Z1Ux7fIo37QKKY6dV26uTR5aKe5VKVllqjS92jzWunhfNfcodYYZbyTlSu9rz67mqN6bQ4oWp52DZy5dPBdl7CvXGo8pF4QwGcCyFmb1dNetdR4HoExj2Gxk/9Lv3lnx/S9dPLfih5esQEoDeWbRqaw3Mfb+rRLEtJSsMlWaXm0ey1XYlSrz0u//2CNm0tUx9l3EzjAx4+1RpuUxSRxVutH1kSuZcdJKkXyLYMn83syKr39gMFPpJj/MeInm3y9/C7cue8OoAfk8sbrSiCuUchVET3dX2Y+0FoaGjc/8YGPZY5bM7yVTMweSlWHe53DqMQfnkrGZowtXIm/FXG0es3pIpQuZlZL2/V+7tp89poytWmPTW8x4e5SleUxrZDTDGjM+JjNOxjOIWm2srWZ1OZ2e4Y0F6XbmvLboWnshsTNAlmsnRJXH+W+LBv/j59pRZs5KNeQxw2UM/QDQ3dVR03NYcMiMXMdBdfO+mqnsVeP5lzePfev6uXbt2IbQtD06+dxflle+Wd9/Fg8NDI5yOy6Nn1dtjzKZx0OX3ZB5z0biPZlxUmuXt5oeUDP3lvrW9fNkhUq1tDWV1xZd7RhBd1cnF59yFEsXz+Xatf2ZCmb6Xl2seMcrRj7QW5e9gd8vfwu7qlAw+QxT2ZQzbSU94yD/czjnursqHlOtabeZyl7fun6e+ePOMemVehuVyOop9uy1R0UlVW0Fvl931ygTsLG7LPV0d7FnVwdnr1xfk9m9WcfaXMmMk1pfbDWDfo1YajVvZbRi1SZ2pe4ZTfJjzKuYK9mck0zfq2vEHJJVafT2dHP/8rew7tPHpVYeeT/G3p5uvnTKUZmKpjvHgGs501apHHmfw+BQ+TdRi8IoquzVMo6Z5cQxbY8pYwKHVnPtrPLYH3od5cgqM2mm2O6uTqSxPR1j92TiJ3YMjbybs1eu59y+DWXvn6RZx9rcXDZOag2eWU0PqNre0nu+/ituvW972fsD7Du1k3269xhjBkkzZ521cj1nrVw/ZoJa3pZch8Shy25gVk93psnDgDmJLv/hL5zG1CljA1+m8cSOyPtrzZbtFdeJyTL/5Fmzpburk2OPmMnSa+7MHIsaHNrFuX0byroTX7hkHr/f9syY95RWdkonjdZq0MtSGGdfvX7UfZLUy905OXlUQEeHGA4KI6/rbtY9nxzcXZayyu5nfrBxzMTKuByUe55ZclUyeZWaYuNylhVNYmAw/Xu4cvVWFhwyI5fZr1ljLLqSGSe1vthqAvtVc2xeBQPw1HPDPPXc2Mi/5QaaSyeo7bVHJ88+X1kJxGMd/QODdHVoTLypNO597Nlc+YjpHxgsG3wyjuK89Lt3jtw79giD9Hd57BEzufmebSNxsiRyBbjMUzlc+YFX5x7vSNre52TY3iuRGXnayJx0mFX2DFi0/KZcZb108qjBiIKJyTOOmdU4SboGZ5XdJ3YMsfSa3XnMM1u+VK4sxRKbvIyol5t8JqVKrdpxPwt5yju21GwKBlzJ1IW0dT4qrQ1STQ+ommPzKpg04g+qUis1OUEtj4IZc/4uo6e7i2lTp4yrZV4N8fP6zA82jlFusUdY/B6zZoTnXZkSosrh/Os31i2KcrICqZVySzhnTTos17vL6oGUVnYPP5lP5nJ561vXP6rHkiQZJ6/cNYZ22cg7qcZzMR6sTz6H0jKbVDArVm3i7JXrU60DWSGE9uzqyHQYyfPOmzlUkKwO3jStxIIFC2zNmjWFXT9rIalS98lq4yTlbaXU2sqNEfnWkxfw++Vvqfl+8fkwfpnzcPEpR7Fkfm/Ze90f5Cmlb10/H786PdxM3vuOh3P7NnDl6q25lHH8/tLKR9+6fs4qE8U6+U5Kzyvnqdfb0z0SzLRaZZykU2KX2UiPMRnw8jM/2FjWay+WYdHymyqW3d4c5bv0eMieJ5Wk1HQW99p3ZIyXdUp84V2vYM2W7Zk95Pi5JBfoK60Lst5PfP3xlkFJa81sQU3ntrqSkXQ88GWgE/gPM1te7viilUxWIa/0ISYV0Xi6veOtsGP330omoZ7uLtafd1zN94tjj5ULdV8vJPjSu44CqFjJlj7v8VSawKgeW94IwaWh7NPs9XnoEPzVMbNHjQ3Nv+CnmZV1soymceiyGzJXE42VU55KvghiGcb7vkqVhID3LJydW8nXwl5dHQztsormY4g8tfI42iRJa+RWS9sqGUmdwP8CfwE8CNwOnGpmv806p1Ylk7fiz/oQiyb+GNIGk6u5Rl7ZuzrFKX+abwGuVqVD5eezFMGiw2bw6/ufyFXh5OXwF05j82PPVny3WfNCqunJVNPomLZHJzueHwZF40LjITZVJa0Dzm4qNSAqMR4l0+ouzK8CNpvZ78zseeA7wEn1vkk1rp+N8kk38g1IV7pGXoaGbVIrGJh4BQPRmFo9FQxEDhR5rvjs88NjynWe0D7J2fLl5v/E+zolTls4m40XHM/vl79l3AoGYOfwMEuvudMVTAaNnJDZ6gP/vUAyJO2DwDH1vkk1s/qXLp5b1iRTNOMZ+G8GqulNOfWntFznGSBPxg0rZ/q876IT6iNkCo8+/Xxh154MNHJCZqv3ZHIh6UxJaySt2batctDGUqqZK9BoT45Wp9Gzk518E2ezju/NeH9Z6RCNWznF0egJma2uZPqB5NTpg0LaKMzsa2a2wMwWzJxZOfx8Kc0arqGRjDesStY1Gz072RldrvOuCRRTy6zz89/20tSoxc5oalHGnVLDg5+2upK5HThc0qGS9gDeDVxf75tU++FMaeD3suiwGakfbIciN+ksqvnGp+3RyXsWzq574XnPwtksmd/LosPyB3rMot6vIF5HPo7oe/EpR3H4C6fV+S4R9ZT9gH32qOr40nJdKaRNadywWiI8L5nfy4p3vmLknKywPOXK6OEvnFZRUZ22cDYXn3JU3SNv14OuTrHv1HS5ROQKv/684zht4ezU8rFXVwddnaP3dHd11sV9eby0tHcZgKQTgIuJXJgvNbPPlTu+aO+ymBefcwM7a3y08Zrt1XiKxd5lFy6ZV3EOTmlUgEWHzeCdC2aPmen+wzsfHjWQWuoSGw0K31UxZlYlpk7p4B/f/vJRz7OayAXJfFz5gVeP/E6GMikl6Un18vN+wlPPjR13iL3LSmdxJ/mLL94yJjLB4S+cxo7nd40qK7B7bsOUDsh6ZKctnM2CQ2aMKWtrtmwfyUunxLQ9OlJljknKnHeeTVY+s2a6V5rbNR6S79BDE4wAAB53SURBVC7+HuLnUuqEEL/30nIfE5+fLLdZIWHSmNIhdpZ4gdRz7DD53EvL05QO8U/vHK0osuqiImf8t60Lcy1I2gZsqfH0/YE/1FGconF5i8XlLRaXt1iqkfcQM6t+rIE2VDLjQdKaWrV5I3B5i8XlLRaXt1gmSt5WH5NxHMdxmhhXMo7jOE5huJKpjq81WoAqcXmLxeUtFpe3WCZEXh+TcZxJhqRbgCvM7D8aLYvjeE/GcWpA0v2SHpM0LZH2/4YK3nGcgCsZx6mdTuCj47mAIvw7dCYtXrgdp3ZWAJ+Q1FO6Q9KfSbpd0pPh/58l9t0i6XOSbgV2AC+SZJI+KOleSU9L+qykwyT9j6SnJF0dologabqkH0raJumJsH3QhOXacarAlYzj1M4a4BbgE8lESTOAG4BLgBcAXwRukPSCxGHvBc4E9mH35ODFwNHAQuAfiAZmTyOKz/cy4NRwXAfwn8AhwGxgEPiXuubMceqEKxnHGR+fBj4sKTkb+i3AvWb2LTPbaWZXAfcAJyaO+aaZbQz74zgonzezp8xsI/Ab4KdhraQngR8D8wHM7HEzu9bMdpjZ08DngD8vOJ+OUxOuZBxnHJjZb4AfAssSybMYG7poC9H6RzEPMJZHE9uDKb/3BpC0l6R/l7RF0lPAL4CesFKs4zQVrmQcZ/ycB3yA3UrkISJTVpLZjF6GYjxzBz4OzAWOMbN9gdeFdI+X7zQdrmQcZ5yY2WZgJfCRkPQj4CWS/krSFEmnAEcS9XjqwT5EPZuBMP5zXp2u6zh1x5WM49SHC4BpEI2ZAG8l6nE8TjSI/1Yzq1eE3ouBbqIIuquBn9Tpuo5Td3zGv+M4jlMY3pNxHMdxCsOVjOM4jlMYrmQcx3GcwnAl4ziO4xTGlEYLMNHsv//+NmfOnEaL4TiO0zKsXbv2D2Y2s/KRY2k7JTNnzhzWrFnTaDEcx3Fqpm9dPytWbeKhgUFm9XSzdPFclszvrXxijUgqjWCRm7ZTMo7jOK1M37p+zrluA4NDwwD0DwxyznUbAApVNLXiYzKO4zgtxIpVm0YUTMzg0DArVm1qkETlcSXjOI7TQjw0MFhVeqNxJeM4jtNCzOrpriq90biScRzHaSGWLp5Ld9foVR26uzpZunhugyQqjw/8O47jtBDx4P5EepeNB1cyjuM4LcaS+b1Nq1RKcXOZ4ziOUxiuZBzHcZzCcCXjOI7jFIYrGcdxHKcwXMk4juM4heFKxnEcxykMVzKO4zhOYbiScRzHcQrDlYzjOI5TGK5kHMdxnMJwJeM4juMUhisZx3EcpzAKUzKSDpZ0s6TfStoo6aMhfYakGyXdG/5PD+mSdImkzZLukvTKxLXOCMffK+mMRPrRkjaEcy6RpKLy4ziO41RPkT2ZncDHzexIYCHwIUlHAsuAn5nZ4cDPwm+ANwOHh78zga9ApJSA84BjgFcB58WKKRzzgcR5xxeYH8dxHKdKCgv1b2YPAw+H7acl3Q30AicBrw+HXQbcAnwypF9uZgasltQj6cBw7I1mth1A0o3A8ZJuAfY1s9Uh/XJgCfDjovLk1I++df0tsx6G4zi1MyHryUiaA8wHbgMOCAoI4BHggLDdCzyQOO3BkFYu/cGU9LT7n0nUO2L27Nm1Z8SpC33r+jnnug0MDg0D0D8wyDnXbQBwReM4k4zCB/4l7Q1cC5xlZk8l94VeixUtg5l9zcwWmNmCmTNnFn07pwIrVm0aUTAxg0PDrFi1qUEStTZ96/pZtPwmDl12A4uW30Tfuv5Gi+Q4IxSqZCR1ESmYK83supD8aDCDEf4/FtL7gYMTpx8U0sqlH5SS7jQ5Dw0MVpXuZBP3CvsHBjF29wpd0TjNQpHeZQK+AdxtZl9M7LoeiD3EzgC+n0g/PXiZLQSeDGa1VcBxkqaHAf/jgFVh31OSFoZ7nZ64ltPEzOrprirdycZ7hU6zU2RPZhHwXuANktaHvxOA5cBfSLoXeFP4DfAj4HfAZuDrwAcBwoD/Z4Hbw98FsRNAOOY/wjn34YP+LcHSxXPp7uocldbd1cnSxXMbJFHr4r1Cp9kp0rvsl0DWvJU3phxvwIcyrnUpcGlK+hrgZeMQ02kA8eC+e5eNn1k93fSnKBTvFTrNwoR4lzlOKUvm97pSqQNLF88d5akH3it0mgtXMo7Twniv0Gl2XMk4TovjvUKnmak48C/pZ3nSHMdxHKeUzJ6MpD2BvYD9g+twPIi/Lxkz6x3HcRwnSTlz2d8CZwGzgLXsVjJPAf9SsFyO4zjOJCBTyZjZl4EvS/qwmf3zBMrkOI7jTBIqDvyb2T9L+jNgTvJ4M7u8QLkcx3GcSUBFJSPpW8BhwHogdsY3wJWM4ziOU5Y8LswLgCPDjHzHcRzHyU2e2GW/Af6kaEEcx3GcyUeensz+wG8l/Rp4Lk40s7cVJpXjOI4zKcijZM4vWgjHcRxncpLHu+znEyGI4ziOM/nI4132NLuXSN4D6AKeNbN9ixTMcRzHaX3y9GT2ibfDCpQnAQuLFMpxHMeZHFS1MqZF9AGLC5LHaXH61vWzaPlNHLrsBhYtv8nXmnecNiePuezkxM8OonkzfyxMIqdl6VvXP2oBrf6BQc65bgOAh6J3nDYlj3fZiYntncD9RCYzxxnFilWbRq3QCDA4NMyKVZtcyThOm5JnTOZ9EyGI0/o8lLLWfLl0x3EmP3nMZQcB/wwsCkn/DXzUzB4sUjCn9ZjV001/ikKZ1dNN37p+XyLY8XLQhuQZ+P9P4HqidWVmAT8IaY4ziqWL59Ld1Tkqrburk2OPmMk5122gf2AQY/dYjTsFtBfxmJ2Xg/Yij5KZaWb/aWY7w983gZkFy+W0IEvm93LRyfPo7elGQG9PNxedPI+b79mWOVbjtA/lxuycyUuegf/HJZ0GXBV+nwo8XpxITiuzZH7vGPPH2SvXpx7rYzXthY/ZtSd5ejJ/A7wLeAR4GHgH4M4ATm5m9XRXle5MTrwctCcVlYyZbTGzt5nZTDN7oZktMbOtEyGcMznIGqtZunhugyRyGoGXg/Ykj3fZocCHGbv8ctlQ/5IuBd4KPGZmLwtpM4CV4Vr3A+8ysydCuJovAycAO4C/NrM7wjlnAOeGy15oZpeF9KOBbwLdwI+IPN58YbUmJDafuVdRe9Oq5aCVPeKaQXZVqpcl3Ql8A9gA7IrTK0VnlvQ64Bng8oSS+Tyw3cyWS1oGTDezT0o6gUiRnQAcA3zZzI4JSmkNUZQBA9YCRwfF9GvgI8BtRErmEjP7caUML1iwwNasWVPpMMdxnDFRLCDqfV108rymVzT1lF3SWjNbUIscecZk/mhml5jZzWb28/iv0klm9gtge0nyScBlYfsyYEki/fIQG2010CPpQKIYaTea2XYzewK4ETg+7NvXzFaH3svliWs5juPUhVb2iGsW2fN4l31Z0nnATxm9MuYdNdzvADN7OGw/AhwQtnuBBxLHPRjSyqU/mJKeiqQzgTMBZs+eXYPYjuO0I63sEdcssudRMvOA9wJvYLe5zMLvmjEzkzQhYyhm9jXgaxCZyybino7jtD7lolg0O80iex5z2TuBF5nZn5vZseGvVgXzaDB1Ef4/FtL7gYMTxx0U0sqlH5SS7kxifBkBZ6Kp1SOuGcpqs3jz5VEyvwF66nS/64EzwvYZwPcT6acrYiHwZDCrrQKOkzRd0nTgOGBV2PeUpIXBM+30xLWcSYiHJHEaxdQpu6vJ6Xt1VRw4b5aymhWBY6IdFvKYy3qAeyTdzu4xGTOzsuH+JV0FvB7YX9KDwHnAcuBqSe8HthBN8oTIO+wEYDORC/P7wk22S/oscHs47gIzi50JPshuF+Yfhz9nkuLLCDgTTZp31sCOIdZs2V62zDVTWU2LwDHR5FEy5yW2BbwWeHelk8zs1Ixdb0w51oAPZVznUuDSlPQ1wMsqyVFPmsHnvJ60Un6aZRDTaQyNKKtpysKAK1dvZcEhM0bdPylf1qBvu5bVPOvJ/FzSfOCviMZnfg98tWjBmo3JtupjWn7OXrmes1aup7fgj7iWCqNZBjGdiafab6+W8pV2TpZSMBjVK0nr8aTRrmU1c0xG0ksknSfpHqL1ZLYSTd481sz+ecIkbBKaxee8XmS10qBYG3Kt9upmGcR0Jp5qvr1aylfWOT17dWWek1RAafKV0s5ltdzA/z1EbspvNbPXBMVS/klOYiabuaaS3OUU6Hg8Z2pV1s0yiOlMPNV8e9WWr751/Zx99frUc8oFQ9mvu2vk/LQedoyX1fLmspOJxl5ulvQT4DtEz6wtmWzmmqz8JEn7iMuZLqByXKrxKOtmGMR0Jp6sshpX9EmqKV996/pZes2dmcrkycEhpk7p4Lmdu8bsGxreNfItZNHb082ty8Y1nXBSkNmTMbM+M3s3cARwM3AW8EJJX5F03EQJ2CxMNnNNWn5KSVOgWS3F86/fOMbkcPbK9cwp6e00W7j3ZpjP4JRn6eK5dHWMbd8++/zOUe+rb10/HUpvB2eV5aFd2d2VWT3dqQomuvdwWTNZK9cN9SZPqP9nzezbZnYi0aTHdcAnC5esyWhlc01aRZrMD4ztomZ9JFktxYHBoVxjPI1Q1lmKpFnmMzjlWTK/l733HGt0GRq2ETNY/C6HU7ol1ZblmGOPKL8AcLnzW6VumAjyuDCPEIJUjoRoaTda0VxTyTMn6SGTxyMnj5ktjdguvnTxXKZO6RiRZ/peXZx34ksL9WTLyn8R8xma2S28GWXLK9PAjqHU8+OK/jM/2Jjaq+iUMiv8SmV55e0PsFdXBzuGxvZmerq7mDZ1Sur5vT3dDX+uzURVSsZpPfJWpHkV6NLFc1PDh+/Z1cETGRVBTFzBJ8/949Au1mzZPqaiiWUfb4VYLv/1duZoZjf3espWL2VVjUzlxkT71vVnlr1dZpmyLV08l6XX3JlpMhsaNro6RFeHRh3T1SHOf9tLAcaUZ1G5B9RuVFxPZrLRbuvJHLrshszJYYKaKom0SgbGfnC10gF0doqh4d2Sd3d18vaje7n5nm1jKrdylV5W/uO8Z7Vka5krtGj5TZkt23IDwBPRw6hVtlLquUZJNTJl3fftR/dy1W0PpJrJsq5Vet3zr9/IwGB2A+niU47KfD/n9m3gytVbR5WxVllvphrGs56M92QmOeUq0uQ4REyeyq5cr6fSB5uHXcCu4dGVxuDQ8KiPOZZ7zZbtXLu2P7M1XK4FnNYri6mlpV9Lz2iiej/16rXVamKMFWn/wCCdUqZSgOgZLFp+U2o5TJbPOS/oHlPBl5I11leq2MuV2XLfxM33bBtzfw93NJo8ATKdFiaPF1mWd1i1g+BL5vcybWpx7Za0j/mq2x4oOy+inKNBqfNDKYNDw3z86jtTHQbSHAlq8ZybqEm+9fLqG48ijZV9OQUDUS+ztBye27dhVGV/7BEz+Z/7tpdVMD3dXakV/bl9Gzh75fpR9yhH2jcRl4Gsc1t1/lwRuLmsDcgTVymLpLkhrTXa092FFA3M7tfdNe5eTL0Q8KVTjuKc6+5iMDFwO22PTj73l2NNGeXMijGxeSbZc4rTLzp5HjDWZJhmOsnzPgT8fvlb8mQ1F+XMTWkmyCxqMbuVq4zzIkY3Mkp/l1L63JNld7wImFJizi2l9HmMxyRaeu6xR8ys6p3VAzeXNTHJArJfokJulHdPJVNFKf0Dg5zbt4Ef3vnwKAUSXyOZ1ggFI5E6mW6/7i4+tnI9pX5Bzz4/zFkr1494ukHUm8jzRAaHhrnytq1j7hf3POJKJWky3LNrtLGgUXGu0sxNxx4xs6ypMY00E2Olwe56tOpL30+59xV7lEF9FFyaLOUUDMCOMIcnHjOs1SSadu4Vq7eO7G8m55IsvCeTg1oVRaUKZSIGCPNUat1dnXUZsK+FSi3SWujqFHtPnVLR263e3L/8LRV7DHkqvIkaOK524D3+Bvbs6hjVO6wkcxEVfTkuPuUooH6OKONheoh/llYW06wEpb2TvM+uksVhvEFvvSdTIKWVRrK1XqkVUSlw3kQMEOYJ3vf8zsZ9iEU0cYaGbdwKptoeH+z+uNPGWJKtz3IUPW8oSVblFadneV6VKpgoLbssH3vEzNz5rwdx5dxoBQPpyiUmthIsOGTGmN7KWSvX87Gr11MmIMEoHkq8s+S14jLcyB6PD/xXIK+iSCOPmaDoAcI816/Q829Lpk6pPkxfubk3efljSgVeFBkRWJB2x/WqxgSalfeb79lWi3g10RvmzUxkz2k8XLF6K5/6XnqPK6+Cgd3zhdKCfcY0Kmq892QqkKfSyCrQeWbHFx2zq9YZ+u1O2izvSvQPDNI7zucde/oVNW8mj9OBWeW4Xmns1901yvU4HqCeqPLX3dXJsUfMZOk1d07I/erFs8+Pr8clYM4LussG+4xphNeb92QqkEcJCFJdfSu5D3d1qvAgesceMbN9Q2c3gDkvGH+jYWBwqK7x1GJ32znLbhjluluOaiujDkUBK5NyX7F664Q2cAaHhrn69geqVo6tjgG33rc9V77LrZFTFN6TqUC5CXsxpSvlxSQ9elI/toK/hb51/Vy7tr/o2zgJbr1ve92vmezdlBvMzROJIU9ZEFFlVM241i4bO4G2ETzfBDI0M43w83LvshzkCT1RaV5DJU+eesaDqtd8AKf5iT3XSl3M4315Ysql0RFcw9urdpj81Dr/yr3LCmbJ/F5WrNpUVsl0SBy67IZMBVHOk+c9X/8Vd2x9sqIf/bl9G0biNHVKnHrMwVy4ZF4uJehMTsp5rg0ODdfsYdVmFqe2oRHrNvmYTA7yeKsMm43Yos9auZ4XnXMD5/ZtGDm/HLfetz1zIbCYc/s2cMXqrSMuicNmXLF6K+/5+q84e+V6VzCO41RkR8lCbxOBm8sq0Leun7NWrh/XPTtUe8vwtIWzU00hjuM4tVDLZN/xmMu8J1OBZG+iVsZjerhi9VZXMI7j1I2Jni/jSqYCXsE7jjPZmEjHIFcyjuM4TmG0vJKRdLykTZI2S1rWaHkcx3Gc3bS0kpHUCfwr8GbgSOBUSUc2VirHcRwnpqWVDPAqYLOZ/c7Mnge+A5zUYJkcx3GcQKsrmV7ggcTvB0PaKCSdKWmNpDXbtk1cRFjHcZx2p9WVTC7M7GtmtsDMFsycmb2Cn+M4jlNfWl3J9AMHJ34fFNIcx3GcJqDVlcztwOGSDpW0B/Bu4Pp63uD+GoLJOY7jNDMTWa+1dIBMM9sp6e+BVUAncKmZjX+KfgmuaBzHcWqj7WKXSdoGbKnx9P2BP9RRnKJxeYvF5S0Wl7dYqpH3EDOraUC77ZTMeJC0ptYgcY3A5S0Wl7dYXN5imSh5W31MxnEcx2liXMk4juM4heFKpjq+1mgBqsTlLRaXt1hc3mKZEHl9TMZxHMcpDO/JOI7jOIXhSiYHjVxOQNLBkm6W9FtJGyV9NKSfL6lf0vrwd0LinHOCrJskLa6UjzCZ9baQvjJMbB2PzPdL2hDkWhPSZki6UdK94f/0kC5Jl4R73yXplYnrnBGOv1fSGYn0o8P1N4dzNQ5Z5yae4XpJT0k6q5mer6RLJT0m6TeJtMKfZ9Y9apR3haR7gkzfk9QT0udIGkw856/WKle5vNcgb+HvX9LU8Htz2D9nHPKuTMh6v6T1zfJ8MTP/K/NHNMnzPuBFwB7AncCRE3j/A4FXhu19gP8lWtbgfOATKccfGWScChwaZO8slw/gauDdYfurwP8Zp8z3A/uXpH0eWBa2lwH/GLZPAH4MCFgI3BbSZwC/C/+nh+3pYd+vw7EK5765ju/6EeCQZnq+wOuAVwK/mcjnmXWPGuU9DpgStv8xIe+c5HEl16lKrqy81yhv4e8f+CDw1bD9bmBlrfKW7P8C8Olmeb7ek6lMQ5cTMLOHzeyOsP00cDcpkaYTnAR8x8yeM7PfA5uJ8pCaj9B6eQPw3XD+ZcCSArJyUrh26T1OAi63iNVAj6QDgcXAjWa23cyeAG4Ejg/79jWz1RaV/MvrKO8bgfvMrNxk3Ql/vmb2C2B7ihxFP8+se1Qtr5n91Mx2hp+rieIMZlKjXFl5r1reMtTz/Sfz8V3gjXFvolZ5w/nvAq4qd42JfL6uZCqTazmBiSB0p+cDt4Wkvw/d1ksTpowsebPSXwAMJCqAeuTPgJ9KWivpzJB2gJk9HLYfAQ6oUd7esF2aXg/ezeiPs1mfL0zM88y6x3j5G6IWccyhktZJ+rmk14a0WuSq97da9PsfOSfsfzIcPx5eCzxqZvcm0hr6fF3JtAiS9gauBc4ys6eArwCHAUcBDxN1kZuF15jZK4lWLP2QpNcld4aWU1O5NQY7+duAa0JSMz/fUUzE86zXPSR9CtgJXBmSHgZmm9l84GPAtyXtO9FypdAy77+EUxndUGr483UlU5mGLycgqYtIwVxpZtcBmNmjZjZsZruArxN118vJm5X+OFG3d0pJes2YWX/4/xjwvSDbo3HXOvx/rEZ5+xltaqnX+3gzcIeZPRpkb9rnG5iI55l1j5qQ9NfAW4H3hMqLYHZ6PGyvJRrXeEmNctXtW52g9z9yTti/Xzi+JsI1TgZWJvLR8OfrSqYyhS8nUI5gY/0GcLeZfTGRnrSF/iUQe5pcD7w7eK4cChxONMCXmo/wsd8MvCOcfwbw/XHIO03SPvE20YDvb4JcsUdT8h7XA6cHz5WFwJOhq74KOE7S9GCqOA5YFfY9JWlheDanj0feBKNagM36fBNMxPPMukfVSDoe+AfgbWa2I5E+U1Jn2H4R0fP8XY1yZeW9Fnkn4v0n8/EO4KZY+dbIm4B7zGzEDNYUz7fUE8D/Ur0wTiDy6roP+NQE3/s1RN3Vu4D14e8E4FvAhpB+PXBg4pxPBVk3kfC8ysoHkUfMr4kGMa8Bpo5D3hcRedbcCWyM70Nka/4ZcC/wX8CMkC7gX4NMG4AFiWv9TZBpM/C+RPoCoo/+PuBfCJOKxyHzNKIW5H6JtKZ5vkTK72FgiMgO/v6JeJ5Z96hR3s1E9vy4DMdeVW8P5WQ9cAdwYq1ylct7DfIW/v6BPcPvzWH/i2qVN6R/E/i7kmMb/nx9xr/jOI5TGG4ucxzHcQrDlYzjOI5TGK5kHMdxnMJwJeM4juMUhisZx3EcpzBcyTiO4ziF4UrGmZRIeoF2hzd/RLvDtj8j6d8KuudZkk4P2xdIelMV5+4l6UpFodd/I+mXIZTQhKIoxP0nyux/q6QLJlImp7XxeTLOpEfS+cAzZvZPBd5jCtFkt1fa7mCI1Zx/DjDTzD4Wfs8F7jez5+oraUU5zqfMswqzw+8AFlli5r7jZOE9GaetkPR6ST8M2+dLukzSf0vaIulkSZ8PvYmfhJhx8eJOP1cUVXqV0sObv4Eo9tnOcM43Jb0jbN8v6TOS7gjXPiLl/ANJxIEys02xgpF0mqRfh57YvyfChBwfrnmnpJ+FtBmS+hRFD14t6eWJvF4q6RZJv5P0kcQz+ZSk/5X0S2BuIv0jihbLu0vSd4JcBtxCFIPMcSriSsZpdw4jUhBvA64AbjazecAg8JagaP4ZeIeZHQ1cCnwu5TqLgLVl7vMHiyJTfwVIM0ddCnxS0q8kXSjpcABJ/w9wClHP4ShgGHiPpJlEgRvfbmavAN4ZrvMZYJ2ZvRz4v0TrhMQcQbSuzKuA8yR1STqaKM7WUURhUf40cfwyYH641t8l0tcQhZR3nIpMqXyI40xqfmxmQ5I2EK1u+JOQvoFoVcG5wMuAGyNLEZ1EcaNKOZBoQbksrgv/1xJFyh2Fma0PAQyPIwp0eLukVxMtpHZ0+A3QTRQVdyHwC4sWzsLM4kWsXkMUrwozuymMTcWh3W8IvaPnJD1GtE7Ia4HvxaYvScngr3cBV0rqA/oS6Y8Bs8rk1XFGcCXjtDvPAZjZLklDtnuQchfR9yFgo5m9usJ1BokCHpa9D1FPJPW7M7NniJTRdZJ2EfUsngcuM7NzksdKOrGCPOVkKCtHgrcQLfV7IvApSfOCOXBPovw6TkXcXOY45dkEzAy9CoKJ6aUpx90NvLjWm0hapLD6YggVfySwhSga7jskvTDsmyHpEKIljF8Xws0jaUa41H8D7wlprycy0z1V5ta/AJZI6la0RMOJ4dwO4GAzuxn4JNFaJ7G320vYHfreccriPRnHKYOZPR8G8C+RtB/RN3MxUfj0JD8mCg9fK4cBXwneWx3ADcC1ZmaSziVazrqDKLz7h8xstaKlra8L6Y8BfwGcD1wq6S5gB7vXBcnK3x2SVhItzfAY0booEJkFrwh5FnCJmQ2EfccC54y5mOOk4C7MjlMnJH0P+Acbvb76pELSAcC3zeyNjZbFaQ1cyThOnQhzWw4ws180WpaikPSnwJCZrW+0LE5r4ErGcRzHKQwf+Hccx3EKw5WM4ziOUxiuZBzHcZzCcCXjOI7jFIYrGcdxHKcw/n9tpGf4795q/wAAAABJRU5ErkJggg==\n",
            "text/plain": [
              "<Figure size 432x288 with 2 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "NbLrzzWR64cv",
        "colab_type": "text"
      },
      "source": [
        "### 標準化資料"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "-I2SZOOT6i4G",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "from sklearn.preprocessing import StandardScaler\n",
        "data = df.drop(['Time'], axis=1)\n",
        "data['Amount'] = StandardScaler().fit_transform(data['Amount'].values.reshape(-1, 1))"
      ],
      "execution_count": 14,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "IiTcJKQ06_mo",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 222
        },
        "outputId": "01f7a88a-56cd-4f9c-fc11-7d2649e9c73b"
      },
      "source": [
        "data.head()"
      ],
      "execution_count": 15,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/html": [
              "<div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>V1</th>\n",
              "      <th>V2</th>\n",
              "      <th>V3</th>\n",
              "      <th>V4</th>\n",
              "      <th>V5</th>\n",
              "      <th>V6</th>\n",
              "      <th>V7</th>\n",
              "      <th>V8</th>\n",
              "      <th>V9</th>\n",
              "      <th>V10</th>\n",
              "      <th>V11</th>\n",
              "      <th>V12</th>\n",
              "      <th>V13</th>\n",
              "      <th>V14</th>\n",
              "      <th>V15</th>\n",
              "      <th>V16</th>\n",
              "      <th>V17</th>\n",
              "      <th>V18</th>\n",
              "      <th>V19</th>\n",
              "      <th>V20</th>\n",
              "      <th>V21</th>\n",
              "      <th>V22</th>\n",
              "      <th>V23</th>\n",
              "      <th>V24</th>\n",
              "      <th>V25</th>\n",
              "      <th>V26</th>\n",
              "      <th>V27</th>\n",
              "      <th>V28</th>\n",
              "      <th>Amount</th>\n",
              "      <th>Class</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>-1.359807</td>\n",
              "      <td>-0.072781</td>\n",
              "      <td>2.536347</td>\n",
              "      <td>1.378155</td>\n",
              "      <td>-0.338321</td>\n",
              "      <td>0.462388</td>\n",
              "      <td>0.239599</td>\n",
              "      <td>0.098698</td>\n",
              "      <td>0.363787</td>\n",
              "      <td>0.090794</td>\n",
              "      <td>-0.551600</td>\n",
              "      <td>-0.617801</td>\n",
              "      <td>-0.991390</td>\n",
              "      <td>-0.311169</td>\n",
              "      <td>1.468177</td>\n",
              "      <td>-0.470401</td>\n",
              "      <td>0.207971</td>\n",
              "      <td>0.025791</td>\n",
              "      <td>0.403993</td>\n",
              "      <td>0.251412</td>\n",
              "      <td>-0.018307</td>\n",
              "      <td>0.277838</td>\n",
              "      <td>-0.110474</td>\n",
              "      <td>0.066928</td>\n",
              "      <td>0.128539</td>\n",
              "      <td>-0.189115</td>\n",
              "      <td>0.133558</td>\n",
              "      <td>-0.021053</td>\n",
              "      <td>0.244964</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>1.191857</td>\n",
              "      <td>0.266151</td>\n",
              "      <td>0.166480</td>\n",
              "      <td>0.448154</td>\n",
              "      <td>0.060018</td>\n",
              "      <td>-0.082361</td>\n",
              "      <td>-0.078803</td>\n",
              "      <td>0.085102</td>\n",
              "      <td>-0.255425</td>\n",
              "      <td>-0.166974</td>\n",
              "      <td>1.612727</td>\n",
              "      <td>1.065235</td>\n",
              "      <td>0.489095</td>\n",
              "      <td>-0.143772</td>\n",
              "      <td>0.635558</td>\n",
              "      <td>0.463917</td>\n",
              "      <td>-0.114805</td>\n",
              "      <td>-0.183361</td>\n",
              "      <td>-0.145783</td>\n",
              "      <td>-0.069083</td>\n",
              "      <td>-0.225775</td>\n",
              "      <td>-0.638672</td>\n",
              "      <td>0.101288</td>\n",
              "      <td>-0.339846</td>\n",
              "      <td>0.167170</td>\n",
              "      <td>0.125895</td>\n",
              "      <td>-0.008983</td>\n",
              "      <td>0.014724</td>\n",
              "      <td>-0.342475</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>-1.358354</td>\n",
              "      <td>-1.340163</td>\n",
              "      <td>1.773209</td>\n",
              "      <td>0.379780</td>\n",
              "      <td>-0.503198</td>\n",
              "      <td>1.800499</td>\n",
              "      <td>0.791461</td>\n",
              "      <td>0.247676</td>\n",
              "      <td>-1.514654</td>\n",
              "      <td>0.207643</td>\n",
              "      <td>0.624501</td>\n",
              "      <td>0.066084</td>\n",
              "      <td>0.717293</td>\n",
              "      <td>-0.165946</td>\n",
              "      <td>2.345865</td>\n",
              "      <td>-2.890083</td>\n",
              "      <td>1.109969</td>\n",
              "      <td>-0.121359</td>\n",
              "      <td>-2.261857</td>\n",
              "      <td>0.524980</td>\n",
              "      <td>0.247998</td>\n",
              "      <td>0.771679</td>\n",
              "      <td>0.909412</td>\n",
              "      <td>-0.689281</td>\n",
              "      <td>-0.327642</td>\n",
              "      <td>-0.139097</td>\n",
              "      <td>-0.055353</td>\n",
              "      <td>-0.059752</td>\n",
              "      <td>1.160686</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>-0.966272</td>\n",
              "      <td>-0.185226</td>\n",
              "      <td>1.792993</td>\n",
              "      <td>-0.863291</td>\n",
              "      <td>-0.010309</td>\n",
              "      <td>1.247203</td>\n",
              "      <td>0.237609</td>\n",
              "      <td>0.377436</td>\n",
              "      <td>-1.387024</td>\n",
              "      <td>-0.054952</td>\n",
              "      <td>-0.226487</td>\n",
              "      <td>0.178228</td>\n",
              "      <td>0.507757</td>\n",
              "      <td>-0.287924</td>\n",
              "      <td>-0.631418</td>\n",
              "      <td>-1.059647</td>\n",
              "      <td>-0.684093</td>\n",
              "      <td>1.965775</td>\n",
              "      <td>-1.232622</td>\n",
              "      <td>-0.208038</td>\n",
              "      <td>-0.108300</td>\n",
              "      <td>0.005274</td>\n",
              "      <td>-0.190321</td>\n",
              "      <td>-1.175575</td>\n",
              "      <td>0.647376</td>\n",
              "      <td>-0.221929</td>\n",
              "      <td>0.062723</td>\n",
              "      <td>0.061458</td>\n",
              "      <td>0.140534</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>-1.158233</td>\n",
              "      <td>0.877737</td>\n",
              "      <td>1.548718</td>\n",
              "      <td>0.403034</td>\n",
              "      <td>-0.407193</td>\n",
              "      <td>0.095921</td>\n",
              "      <td>0.592941</td>\n",
              "      <td>-0.270533</td>\n",
              "      <td>0.817739</td>\n",
              "      <td>0.753074</td>\n",
              "      <td>-0.822843</td>\n",
              "      <td>0.538196</td>\n",
              "      <td>1.345852</td>\n",
              "      <td>-1.119670</td>\n",
              "      <td>0.175121</td>\n",
              "      <td>-0.451449</td>\n",
              "      <td>-0.237033</td>\n",
              "      <td>-0.038195</td>\n",
              "      <td>0.803487</td>\n",
              "      <td>0.408542</td>\n",
              "      <td>-0.009431</td>\n",
              "      <td>0.798278</td>\n",
              "      <td>-0.137458</td>\n",
              "      <td>0.141267</td>\n",
              "      <td>-0.206010</td>\n",
              "      <td>0.502292</td>\n",
              "      <td>0.219422</td>\n",
              "      <td>0.215153</td>\n",
              "      <td>-0.073403</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "         V1        V2        V3        V4  ...       V27       V28    Amount  Class\n",
              "0 -1.359807 -0.072781  2.536347  1.378155  ...  0.133558 -0.021053  0.244964      0\n",
              "1  1.191857  0.266151  0.166480  0.448154  ... -0.008983  0.014724 -0.342475      0\n",
              "2 -1.358354 -1.340163  1.773209  0.379780  ... -0.055353 -0.059752  1.160686      0\n",
              "3 -0.966272 -0.185226  1.792993 -0.863291  ...  0.062723  0.061458  0.140534      0\n",
              "4 -1.158233  0.877737  1.548718  0.403034  ...  0.219422  0.215153 -0.073403      0\n",
              "\n",
              "[5 rows x 30 columns]"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 15
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "f3v8WwV17Mxn",
        "colab_type": "text"
      },
      "source": [
        "### 將資料分為訓練與測試資料集"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "uGuLPkTn7ad0",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "X = data.iloc[:,:-1]\n",
        "y = data.iloc[:,-1]"
      ],
      "execution_count": 16,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "LP71XCOj7AvQ",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "from sklearn.model_selection import train_test_split\n",
        "X_train, X_test, y_train, y_test = train_test_split(X,y, test_size=0.2, random_state=42)"
      ],
      "execution_count": 18,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "az0bKwaB7i6c",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 35
        },
        "outputId": "d93273f3-fde4-44a3-8c1f-3e29652d8219"
      },
      "source": [
        "X_train.shape"
      ],
      "execution_count": 19,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "(227845, 29)"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 19
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "yz2WKeL87omZ",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 35
        },
        "outputId": "62a2c1b2-850a-451e-d30d-7ff16fd92692"
      },
      "source": [
        "X_test.shape"
      ],
      "execution_count": 20,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "(56962, 29)"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 20
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "w7g7z6mL7vv6",
        "colab_type": "text"
      },
      "source": [
        "### 跑 ANN"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "UtkZ0JHv7re4",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "% tensorflow_version 2.x"
      ],
      "execution_count": 21,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "jYYT1Aux71K4",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "import tensorflow\n",
        "import tensorflow as tf\n"
      ],
      "execution_count": 24,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "6mqVRnLR7-H2",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "from tensorflow.keras.models import Sequential\n",
        "from tensorflow.keras.layers import Dense, Dropout\n",
        "model = Sequential()\n",
        "model.add(Dense(14, activation = 'relu', input_shape=(29,)))\n",
        "model.add(Dropout(rate=0.1))\n",
        "model.add(Dense(14, activation = 'relu'))\n",
        "model.add(Dropout(rate=0.1))\n",
        "model.add(Dense(1, activation='sigmoid'))"
      ],
      "execution_count": 34,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "zIRmCS338nmX",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 329
        },
        "outputId": "40154e26-1bf4-4977-8646-1fd0c4297d30"
      },
      "source": [
        "model.summary()"
      ],
      "execution_count": 35,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Model: \"sequential_4\"\n",
            "_________________________________________________________________\n",
            "Layer (type)                 Output Shape              Param #   \n",
            "=================================================================\n",
            "dense_6 (Dense)              (None, 14)                420       \n",
            "_________________________________________________________________\n",
            "dropout_2 (Dropout)          (None, 14)                0         \n",
            "_________________________________________________________________\n",
            "dense_7 (Dense)              (None, 14)                210       \n",
            "_________________________________________________________________\n",
            "dropout_3 (Dropout)          (None, 14)                0         \n",
            "_________________________________________________________________\n",
            "dense_8 (Dense)              (None, 1)                 15        \n",
            "=================================================================\n",
            "Total params: 645\n",
            "Trainable params: 645\n",
            "Non-trainable params: 0\n",
            "_________________________________________________________________\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "uCJP6GOI8uF_",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "model.compile(optimizer='adam', metrics = ['accuracy'], loss = 'binary_crossentropy')"
      ],
      "execution_count": 36,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "GiEPRgxq9A65",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "model.fit(X_train, y_train, \n",
        "    epochs = 100,\n",
        "    validation_data=(X_test, y_test))"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "TwEWx6Qy9Lny",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "y_pred = (model.predict(X_test) > 0.5).flatten().astype(int)"
      ],
      "execution_count": 43,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "5s_sMGTC9uAi",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 35
        },
        "outputId": "63f0f631-e973-4047-ae80-bb3fd40b01ff"
      },
      "source": [
        "from sklearn.metrics import accuracy_score, confusion_matrix\n",
        "accuracy_score(y_test, y_pred)"
      ],
      "execution_count": 45,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "0.999385555282469"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 45
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "MQcLoCCF9-2B",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 52
        },
        "outputId": "1e7a639d-0674-41ca-bf2e-80347a346fe8"
      },
      "source": [
        "confusion_matrix(y_test, y_pred)"
      ],
      "execution_count": 46,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "array([[56851,    13],\n",
              "       [   22,    76]])"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 46
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "qIyKn4m6-KC4",
        "colab_type": "text"
      },
      "source": [
        "### Autoencoders"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "kBPJcB67-Fug",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "X_train = X_train[y_train ==0]"
      ],
      "execution_count": 48,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "4WqH0BwM-ZSs",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "\n",
        "input_dim = X_train.shape[1]"
      ],
      "execution_count": 49,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "TZxiHiZU-oyL",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 35
        },
        "outputId": "69742b0b-6fa0-4e21-c59b-8eb7ebf991d9"
      },
      "source": [
        "input_dim"
      ],
      "execution_count": 50,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "29"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 50
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "cogkFNHh-pzK",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "encoding_dim = 14"
      ],
      "execution_count": 51,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "cK_TAaE_--dS",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "from tensorflow.keras.layers import Input\n",
        "from tensorflow.keras import Model\n",
        "from tensorflow.keras import regularizers\n",
        "input_layer = Input(shape=(input_dim, ))\n",
        "\n",
        "encoder = Dense(encoding_dim, activation=\"tanh\", \n",
        "                activity_regularizer=regularizers.l1(10e-5))(input_layer)\n",
        "encoder = Dense(7, activation=\"relu\")(encoder)\n",
        "\n",
        "decoder = Dense(7, activation='tanh')(encoder)\n",
        "decoder = Dense(input_dim, activation='relu')(decoder)\n",
        "\n",
        "autoencoder = Model(inputs=input_layer, outputs=decoder)"
      ],
      "execution_count": 56,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "TO8ymzAx_gmF",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 384
        },
        "outputId": "68ef06fd-7262-4fda-a78d-257bd9c1daa4"
      },
      "source": [
        "from keras.callbacks import ModelCheckpoint\n",
        "nb_epoch = 10\n",
        "batch_size = 32\n",
        "\n",
        "autoencoder.compile(optimizer='adam', \n",
        "                    loss='mean_squared_error', \n",
        "                    metrics=['accuracy'])\n",
        "\n",
        "checkpointer = ModelCheckpoint(filepath=\"model.h5\",\n",
        "                               verbose=0,\n",
        "                               save_best_only=True)\n",
        "\n",
        "history = autoencoder.fit(X_train, X_train,\n",
        "                    epochs=nb_epoch,\n",
        "                    batch_size=batch_size,\n",
        "                    shuffle=True,\n",
        "                    validation_data=(X_test, X_test),\n",
        "                    verbose=1,\n",
        "                    callbacks=[checkpointer]).history"
      ],
      "execution_count": 59,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Epoch 1/10\n",
            "7108/7108 [==============================] - 11s 2ms/step - loss: 0.8223 - accuracy: 0.5735 - val_loss: 0.7787 - val_accuracy: 0.6792\n",
            "Epoch 2/10\n",
            "7108/7108 [==============================] - 11s 2ms/step - loss: 0.7371 - accuracy: 0.6904 - val_loss: 0.7590 - val_accuracy: 0.6981\n",
            "Epoch 3/10\n",
            "7108/7108 [==============================] - 11s 2ms/step - loss: 0.7240 - accuracy: 0.6964 - val_loss: 0.7510 - val_accuracy: 0.6993\n",
            "Epoch 4/10\n",
            "7108/7108 [==============================] - 11s 2ms/step - loss: 0.7170 - accuracy: 0.7030 - val_loss: 0.7476 - val_accuracy: 0.7042\n",
            "Epoch 5/10\n",
            "7108/7108 [==============================] - 11s 2ms/step - loss: 0.7134 - accuracy: 0.7074 - val_loss: 0.7434 - val_accuracy: 0.7140\n",
            "Epoch 6/10\n",
            "7108/7108 [==============================] - 11s 2ms/step - loss: 0.7115 - accuracy: 0.7097 - val_loss: 0.7433 - val_accuracy: 0.7188\n",
            "Epoch 7/10\n",
            "7108/7108 [==============================] - 11s 2ms/step - loss: 0.7094 - accuracy: 0.7135 - val_loss: 0.7409 - val_accuracy: 0.7179\n",
            "Epoch 8/10\n",
            "7108/7108 [==============================] - 11s 2ms/step - loss: 0.7078 - accuracy: 0.7157 - val_loss: 0.7402 - val_accuracy: 0.7180\n",
            "Epoch 9/10\n",
            "7108/7108 [==============================] - 11s 2ms/step - loss: 0.7062 - accuracy: 0.7172 - val_loss: 0.7404 - val_accuracy: 0.7176\n",
            "Epoch 10/10\n",
            "7108/7108 [==============================] - 11s 2ms/step - loss: 0.7044 - accuracy: 0.7200 - val_loss: 0.7374 - val_accuracy: 0.7248\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "sKBUo-G6ANgX",
        "colab_type": "text"
      },
      "source": [
        "### 評估模型"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "eePvarLl_u-j",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 313
        },
        "outputId": "2d19eeb7-c4ba-42ee-af50-22b7b00dac34"
      },
      "source": [
        "plt.plot(history['loss'])\n",
        "plt.plot(history['val_loss'])\n",
        "plt.title('model loss')\n",
        "plt.ylabel('loss')\n",
        "plt.xlabel('epoch')\n",
        "plt.legend(['train', 'test'], loc='upper right')"
      ],
      "execution_count": 60,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<matplotlib.legend.Legend at 0x7f01c851a6d8>"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 60
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "kywMWKcGAQar",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "predictions = autoencoder.predict(X_test)"
      ],
      "execution_count": 61,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "BZhh8mEMAT97",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 35
        },
        "outputId": "fc63e0fa-5085-461f-d021-20ae4117286a"
      },
      "source": [
        "predictions.shape"
      ],
      "execution_count": 65,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "(56962, 29)"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 65
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "DMo53ViLAVcK",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "import numpy as np\n",
        "mse = np.mean(np.power(X_test - predictions, 2), axis=1)\n",
        "error_df = pandas.DataFrame({'reconstruction_error': mse,\n",
        "                        'true_class': y_test})"
      ],
      "execution_count": 67,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Oxq4cISEAg94",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 415
        },
        "outputId": "16e496e0-a4d6-45f1-ccc0-51d9d3e9b5ca"
      },
      "source": [
        "error_df"
      ],
      "execution_count": 68,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/html": [
              "<div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>reconstruction_error</th>\n",
              "      <th>true_class</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>43428</th>\n",
              "      <td>80.774561</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>49906</th>\n",
              "      <td>0.754743</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>29474</th>\n",
              "      <td>0.229801</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>276481</th>\n",
              "      <td>0.378687</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>278846</th>\n",
              "      <td>0.171890</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>...</th>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>75723</th>\n",
              "      <td>0.639550</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>252263</th>\n",
              "      <td>0.203675</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>221246</th>\n",
              "      <td>0.531668</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>81910</th>\n",
              "      <td>0.152339</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>59490</th>\n",
              "      <td>1.641981</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "<p>56962 rows × 2 columns</p>\n",
              "</div>"
            ],
            "text/plain": [
              "        reconstruction_error  true_class\n",
              "43428              80.774561           1\n",
              "49906               0.754743           0\n",
              "29474               0.229801           0\n",
              "276481              0.378687           0\n",
              "278846              0.171890           0\n",
              "...                      ...         ...\n",
              "75723               0.639550           0\n",
              "252263              0.203675           0\n",
              "221246              0.531668           0\n",
              "81910               0.152339           0\n",
              "59490               1.641981           0\n",
              "\n",
              "[56962 rows x 2 columns]"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 68
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "mzUeNiinAlyo",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "from sklearn.metrics import (confusion_matrix, precision_recall_curve, auc,\n",
        "                             roc_curve, recall_score, classification_report, f1_score,\n",
        "                             precision_recall_fscore_support)"
      ],
      "execution_count": 69,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "LXskdAfSA3qG",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 295
        },
        "outputId": "c3339383-7095-4936-f268-365fbd936f8f"
      },
      "source": [
        "fpr, tpr, thresholds = roc_curve(error_df.true_class, error_df.reconstruction_error)\n",
        "roc_auc = auc(fpr, tpr)\n",
        "\n",
        "plt.title('Receiver Operating Characteristic')\n",
        "plt.plot(fpr, tpr, label='AUC = %0.4f'% roc_auc)\n",
        "plt.legend(loc='lower right')\n",
        "plt.plot([0,1],[0,1],'r--')\n",
        "plt.xlim([-0.001, 1])\n",
        "plt.ylim([0, 1.001])\n",
        "plt.ylabel('True Positive Rate')\n",
        "plt.xlabel('False Positive Rate')\n",
        "plt.show()"
      ],
      "execution_count": 70,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "EtBJP1T6A7Bj",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 52
        },
        "outputId": "6a5c0dbe-58b1-40ee-f54c-33e76b4b2000"
      },
      "source": [
        "threshold = 1.063395\n",
        "y_pred = [1 if e > threshold else 0 for e in error_df.reconstruction_error.values]\n",
        "conf_matrix = confusion_matrix(error_df.true_class, y_pred)\n",
        "conf_matrix"
      ],
      "execution_count": 134,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "array([[52136,  4728],\n",
              "       [    8,    90]])"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 134
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "pgnc3r0OBSLi",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 295
        },
        "outputId": "66400237-880b-495c-a880-aeda2ca7760f"
      },
      "source": [
        "groups = error_df.groupby('true_class')\n",
        "fig, ax = plt.subplots()\n",
        "\n",
        "for name, group in groups:\n",
        "    ax.plot(group.index, group.reconstruction_error, marker='o', ms=3.5, linestyle='',\n",
        "            label= \"Fraud\" if name == 1 else \"Normal\")\n",
        "ax.hlines(threshold, ax.get_xlim()[0], ax.get_xlim()[1], colors=\"r\", zorder=100, label='Threshold')\n",
        "ax.legend()\n",
        "plt.title(\"Reconstruction error for different classes\")\n",
        "plt.ylabel(\"Reconstruction error\")\n",
        "plt.xlabel(\"Data point index\")\n",
        "plt.show()"
      ],
      "execution_count": 72,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "fzrX-MY7BvJ-",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "i = np.arange(len(tpr)) \n",
        "roc = pandas.DataFrame(\n",
        "    {'fpr'   : pandas.Series(fpr, index=i),\n",
        "     'tpr'   : pandas.Series(tpr, index = i), \n",
        "     '1-fpr' : pandas.Series(1-fpr, index = i), \n",
        "     'tf'    : pandas.Series(tpr - (1-fpr), index = i), \n",
        "     'thresholds' : pandas.Series(thresholds, index = i)})"
      ],
      "execution_count": 130,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "yfBtGdoNX-L3",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 225
        },
        "outputId": "ea2cdd13-e6f9-487c-e722-515df1119319"
      },
      "source": [
        "roc['tf'].abs().argsort()"
      ],
      "execution_count": 132,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "0        157\n",
              "1        158\n",
              "2        156\n",
              "3        155\n",
              "4        159\n",
              "        ... \n",
              "1008    1009\n",
              "1009    1010\n",
              "1010    1011\n",
              "1011       0\n",
              "1012    1012\n",
              "Name: tf, Length: 1013, dtype: int64"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 132
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "BTZv5Fz8DUcV",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 121
        },
        "outputId": "1eaf3c0d-cc37-4ec9-b973-0838ff110e93"
      },
      "source": [
        "roc.loc[(roc.tf-0).abs().argsort()[0]]"
      ],
      "execution_count": 133,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "fpr           0.083163\n",
              "tpr           0.918367\n",
              "1-fpr         0.916837\n",
              "tf            0.001531\n",
              "thresholds    1.063395\n",
              "Name: 157, dtype: float64"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 133
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "sa7FANLGYaNs",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 283
        },
        "outputId": "c36c5a32-0ae8-4091-93fa-761493a6ee55"
      },
      "source": [
        "plt.plot(range(len(roc)), roc['tpr']);\n",
        "plt.plot(range(len(roc)), roc['1-fpr'])"
      ],
      "execution_count": 141,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "[<matplotlib.lines.Line2D at 0x7f01bf3d3b00>]"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 141
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "QQNbN7ImKiEl",
        "colab_type": "text"
      },
      "source": [
        "## Leaky ReLU"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "HkkbtKE5KmqZ",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 35
        },
        "outputId": "4178f236-cfbb-4768-a2fe-98a29488964f"
      },
      "source": [
        "def relu_function(x):\n",
        "  return np.maximum(0,x)\n",
        "a = np.array([-1,1,2])\n",
        "relu_function(a)"
      ],
      "execution_count": 82,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "array([0, 1, 2])"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 82
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "9oS4liN_LHFU",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 283
        },
        "outputId": "efade516-59e8-4e83-ab40-4be1bc9dbf43"
      },
      "source": [
        "x = np.arange(-10,10,0.1)\n",
        "plt.plot(x, relu_function(x))"
      ],
      "execution_count": 83,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "[<matplotlib.lines.Line2D at 0x7f01c86ea898>]"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 83
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Y2RSXsumLQJU",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 35
        },
        "outputId": "42ffff4c-0624-4a4e-b650-a96db1dc10c2"
      },
      "source": [
        "def leaky_relu(z, alpha =0.01):\n",
        "  return np.maximum(z, alpha * z)\n",
        "\n",
        "leaky_relu(a)"
      ],
      "execution_count": 84,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "array([-0.01,  1.  ,  2.  ])"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 84
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "tq4j7VYsN9Yo",
        "colab_type": "text"
      },
      "source": [
        "## Over saturation"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "WPpwgoZcOAQK",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 283
        },
        "outputId": "eb70f81c-a721-4c96-9195-3d2dd92682a1"
      },
      "source": [
        "def sigmoid(x):\n",
        "  return 1 / (1 + np.exp(-x))\n",
        "x = np.arange(-10,10,0.1)\n",
        "plt.plot(x, sigmoid(x))"
      ],
      "execution_count": 90,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "[<matplotlib.lines.Line2D at 0x7f01c643e8d0>]"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 90
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "8xGsej5nOPx5",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 107
        },
        "outputId": "c6d67392-a59b-4c27-f052-e53b89c680d9"
      },
      "source": [
        "x1 = np.array([-1000,-100,-10,-1,1,10,100, 1000])\n",
        "sigmoid(x1)\n",
        "\n",
        "sigmoid((x1 - x1.mean()) / x1.std())"
      ],
      "execution_count": 94,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.6/dist-packages/ipykernel_launcher.py:2: RuntimeWarning: overflow encountered in exp\n",
            "  \n"
          ],
          "name": "stderr"
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "array([0.12025952, 0.45041415, 0.49502523, 0.49950251, 0.50049749,\n",
              "       0.50497477, 0.54958585, 0.87974048])"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 94
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "A_JdMFZKJ4lJ",
        "colab_type": "text"
      },
      "source": [
        "## GAN"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "j2riYNXTDesQ",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "from tensorflow.keras.datasets import mnist\n",
        "from tensorflow.keras.layers import Input, Dense, Reshape, Flatten, Dropout\n",
        "from tensorflow.keras.layers import BatchNormalization, Activation, ZeroPadding2D\n",
        "from tensorflow.keras.layers import LeakyReLU\n",
        "from tensorflow.keras.layers import UpSampling2D, Conv2D\n",
        "from tensorflow.keras.models import Sequential, Model\n",
        "from tensorflow.keras.optimizers import Adam\n",
        "\n",
        "import matplotlib.pyplot as plt\n",
        "\n",
        "import sys\n",
        "\n",
        "import numpy as np"
      ],
      "execution_count": 17,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "JkF80h6KKNOT",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "img_rows = 28\n",
        "img_cols = 28\n",
        "channels = 1\n",
        "img_shape = (img_rows, img_cols, channels)\n",
        "\n",
        "latent_dim = 100"
      ],
      "execution_count": 81,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "3TZCTzQyKSxy",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "def build_discriminator():\n",
        "\n",
        "    model = Sequential()\n",
        "\n",
        "    model.add(Flatten(input_shape=img_shape))\n",
        "    model.add(Dense(512))\n",
        "    model.add(LeakyReLU(alpha=0.2))\n",
        "    model.add(Dense(256))\n",
        "    model.add(LeakyReLU(alpha=0.2))\n",
        "    model.add(Dense(1, activation='sigmoid'))\n",
        "    model.summary()\n",
        "\n",
        "    img = Input(shape=img_shape)\n",
        "    validity = model(img)\n",
        "\n",
        "    return Model(img, validity)"
      ],
      "execution_count": 85,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "gu0DkjWtL9f2",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 364
        },
        "outputId": "af6d8345-38fa-499c-c2a8-2a9e60bfe68c"
      },
      "source": [
        "optimizer = Adam(0.0002, 0.5)\n",
        "\n",
        "# Build and compile the discriminator\n",
        "discriminator = build_discriminator()"
      ],
      "execution_count": 86,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Model: \"sequential_5\"\n",
            "_________________________________________________________________\n",
            "Layer (type)                 Output Shape              Param #   \n",
            "=================================================================\n",
            "flatten (Flatten)            (None, 784)               0         \n",
            "_________________________________________________________________\n",
            "dense_17 (Dense)             (None, 512)               401920    \n",
            "_________________________________________________________________\n",
            "leaky_re_lu (LeakyReLU)      (None, 512)               0         \n",
            "_________________________________________________________________\n",
            "dense_18 (Dense)             (None, 256)               131328    \n",
            "_________________________________________________________________\n",
            "leaky_re_lu_1 (LeakyReLU)    (None, 256)               0         \n",
            "_________________________________________________________________\n",
            "dense_19 (Dense)             (None, 1)                 257       \n",
            "=================================================================\n",
            "Total params: 533,505\n",
            "Trainable params: 533,505\n",
            "Non-trainable params: 0\n",
            "_________________________________________________________________\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "jsmcEU_ZMjXD",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "discriminator.compile(loss='binary_crossentropy',\n",
        "    optimizer=optimizer,\n",
        "    metrics=['accuracy'])"
      ],
      "execution_count": 87,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "48cutWTUMq8i",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "def build_generator():\n",
        "\n",
        "    model = Sequential()\n",
        "\n",
        "    model.add(Dense(256, input_dim=latent_dim))\n",
        "    model.add(LeakyReLU(alpha=0.2))\n",
        "    model.add(BatchNormalization())\n",
        "    model.add(Dense(512))\n",
        "    model.add(LeakyReLU(alpha=0.2))\n",
        "    model.add(BatchNormalization())\n",
        "    model.add(Dense(1024))\n",
        "    model.add(LeakyReLU(alpha=0.2))\n",
        "    model.add(BatchNormalization())\n",
        "    model.add(Dense(np.prod(img_shape), activation='tanh'))\n",
        "    model.add(Reshape(img_shape))\n",
        "\n",
        "    model.summary()\n",
        "\n",
        "    noise = Input(shape=(latent_dim,))\n",
        "    img = model(noise)\n",
        "\n",
        "    return Model(noise, img)"
      ],
      "execution_count": 95,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "88mJg354M09B",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 537
        },
        "outputId": "079e4bf5-19c3-4e41-8ee2-d6688d027fd3"
      },
      "source": [
        "# Build the generator\n",
        "generator = build_generator()"
      ],
      "execution_count": 96,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Model: \"sequential_6\"\n",
            "_________________________________________________________________\n",
            "Layer (type)                 Output Shape              Param #   \n",
            "=================================================================\n",
            "dense_20 (Dense)             (None, 256)               25856     \n",
            "_________________________________________________________________\n",
            "leaky_re_lu_2 (LeakyReLU)    (None, 256)               0         \n",
            "_________________________________________________________________\n",
            "batch_normalization (BatchNo (None, 256)               1024      \n",
            "_________________________________________________________________\n",
            "dense_21 (Dense)             (None, 512)               131584    \n",
            "_________________________________________________________________\n",
            "leaky_re_lu_3 (LeakyReLU)    (None, 512)               0         \n",
            "_________________________________________________________________\n",
            "batch_normalization_1 (Batch (None, 512)               2048      \n",
            "_________________________________________________________________\n",
            "dense_22 (Dense)             (None, 1024)              525312    \n",
            "_________________________________________________________________\n",
            "leaky_re_lu_4 (LeakyReLU)    (None, 1024)              0         \n",
            "_________________________________________________________________\n",
            "batch_normalization_2 (Batch (None, 1024)              4096      \n",
            "_________________________________________________________________\n",
            "dense_23 (Dense)             (None, 784)               803600    \n",
            "_________________________________________________________________\n",
            "reshape (Reshape)            (None, 28, 28, 1)         0         \n",
            "=================================================================\n",
            "Total params: 1,493,520\n",
            "Trainable params: 1,489,936\n",
            "Non-trainable params: 3,584\n",
            "_________________________________________________________________\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "kGUCkFbqPUJV",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 35
        },
        "outputId": "33aa3cc3-4b28-4f25-80d4-1c9cd43ddc91"
      },
      "source": [
        "z = Input(shape=(latent_dim,))\n",
        "img = generator(z)\n",
        "img.shape"
      ],
      "execution_count": 97,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "TensorShape([None, 28, 28, 1])"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 97
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "TpG4zsg2PWLE",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "# For the combined model we will only train the generator\n",
        "discriminator.trainable = False\n",
        "\n",
        "# The discriminator takes generated images as input and determines validity\n",
        "validity = discriminator(img)\n",
        "\n",
        "# The combined model  (stacked generator and discriminator)\n",
        "# Trains the generator to fool the discriminator\n",
        "combined = Model(z, validity)\n",
        "combined.compile(loss='binary_crossentropy', optimizer=optimizer)"
      ],
      "execution_count": 98,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "381BN892P8KE",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "def sample_images(epoch):\n",
        "    r, c = 5, 5\n",
        "    noise = np.random.normal(0, 1, (r * c, latent_dim))\n",
        "    gen_imgs = generator.predict(noise)\n",
        "\n",
        "    # Rescale images 0 - 1\n",
        "    gen_imgs = 0.5 * gen_imgs + 0.5\n",
        "\n",
        "    fig, axs = plt.subplots(r, c)\n",
        "    cnt = 0\n",
        "    for i in range(r):\n",
        "        for j in range(c):\n",
        "            axs[i,j].imshow(gen_imgs[cnt, :,:,0], cmap='gray')\n",
        "            axs[i,j].axis('off')\n",
        "            cnt += 1\n",
        "    fig.savefig(\"/content/images/%d.png\" % epoch)\n",
        "    plt.close()"
      ],
      "execution_count": 102,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "dhcEgHzWPy9p",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "def train(epochs, batch_size=128, sample_interval=50):\n",
        "\n",
        "    # Load the dataset\n",
        "    (X_train, _), (_, _) = mnist.load_data()\n",
        "\n",
        "    # Rescale -1 to 1\n",
        "    X_train = X_train / 127.5 - 1.\n",
        "    X_train = np.expand_dims(X_train, axis=3)\n",
        "\n",
        "    # Adversarial ground truths\n",
        "    valid = np.ones((batch_size, 1))\n",
        "    fake = np.zeros((batch_size, 1))\n",
        "\n",
        "    for epoch in range(epochs):\n",
        "\n",
        "        # ---------------------\n",
        "        #  Train Discriminator\n",
        "        # ---------------------\n",
        "\n",
        "        # Select a random batch of images\n",
        "        idx = np.random.randint(0, X_train.shape[0], batch_size)\n",
        "        imgs = X_train[idx]\n",
        "\n",
        "        noise = np.random.normal(0, 1, (batch_size, latent_dim))\n",
        "\n",
        "        # Generate a batch of new images\n",
        "        gen_imgs = generator.predict(noise)\n",
        "\n",
        "        # Train the discriminator\n",
        "        d_loss_real = discriminator.train_on_batch(imgs, valid)\n",
        "        d_loss_fake = discriminator.train_on_batch(gen_imgs, fake)\n",
        "        d_loss = 0.5 * np.add(d_loss_real, d_loss_fake)\n",
        "\n",
        "        # ---------------------\n",
        "        #  Train Generator\n",
        "        # ---------------------\n",
        "\n",
        "        noise = np.random.normal(0, 1, (batch_size, latent_dim))\n",
        "\n",
        "        # Train the generator (to have the discriminator label samples as valid)\n",
        "        g_loss = combined.train_on_batch(noise, valid)\n",
        "\n",
        "        # Plot the progress\n",
        "        print (\"%d [D loss: %f, acc.: %.2f%%] [G loss: %f]\" % (epoch, d_loss[0], 100*d_loss[1], g_loss))\n",
        "\n",
        "        # If at save interval => save generated image samples\n",
        "        if epoch % sample_interval == 0:\n",
        "            sample_images(epoch)\n"
      ],
      "execution_count": 103,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "PtmgCJ5nQjqu",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "! rm -rf /content/images\n",
        "! mkdir /content/images"
      ],
      "execution_count": 104,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "JIX5XbV1Qmdb",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "\n",
        "train(1000, batch_size=128, sample_interval=50)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "irUKCH9FQoQc",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "n1 = np.random.normal(0, 1, (1, latent_dim))"
      ],
      "execution_count": 125,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Wco15zECSsKE",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "#n1"
      ],
      "execution_count": 126,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "TjJ9HOjoSspR",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "gen_imgs = generator.predict(n1)"
      ],
      "execution_count": 127,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "CorjmvFpSx9x",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "gen_imgs = 0.5 * gen_imgs + 0.5\n",
        "\n",
        "   "
      ],
      "execution_count": 128,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "8KjXTJw1TDRv",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 283
        },
        "outputId": "01028dbb-1b58-43a7-c29b-3434a1375bb0"
      },
      "source": [
        "\n",
        "plt.imshow(gen_imgs[0, :,:,0], cmap='gray')"
      ],
      "execution_count": 129,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<matplotlib.image.AxesImage at 0x7f01bf4352b0>"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 129
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "klMWcQOqaL6C",
        "colab_type": "text"
      },
      "source": [
        "## CelebA"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "CdxjCWIZaXw-",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 52
        },
        "outputId": "7f892164-e9c2-4aea-e15e-db86322ef760"
      },
      "source": [
        "! gdown https://drive.google.com/uc?id=0B7EVK8r0v71pZjFTYXZWM3FlRnM"
      ],
      "execution_count": 2,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Permission denied: https://drive.google.com/uc?id=0B7EVK8r0v71pZjFTYXZWM3FlRnM\n",
            "Maybe you need to change permission over 'Anyone with the link'?\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "LXjOL325agru",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "! unzip img_align_celeba.zip > 1.log"
      ],
      "execution_count": 143,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "YZxNls_jtgZE",
        "colab_type": "text"
      },
      "source": [
        "## DCGAN"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "rjIwE4eCalmL",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "\n",
        "from tensorflow.keras.datasets import mnist\n",
        "from tensorflow.keras.layers import Input, Dense, Reshape, Flatten, Dropout\n",
        "from tensorflow.keras.layers import BatchNormalization, Activation, ZeroPadding2D\n",
        "from tensorflow.keras.layers import UpSampling2D, Conv2D\n",
        "from tensorflow.keras.models import Sequential, Model\n",
        "from tensorflow.keras.optimizers import Adam\n",
        "\n",
        "import matplotlib.pyplot as plt\n",
        "\n",
        "import sys\n",
        "\n",
        "import numpy as np"
      ],
      "execution_count": 10,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "QEB68NA6thr3",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "class DCGAN():\n",
        "    def __init__(self):\n",
        "        # Input shape\n",
        "        self.img_rows = 28\n",
        "        self.img_cols = 28\n",
        "        self.channels = 1\n",
        "        self.img_shape = (self.img_rows, self.img_cols, self.channels)\n",
        "        self.latent_dim = 100\n",
        "\n",
        "        optimizer = Adam(0.0002, 0.5)\n",
        "\n",
        "        # Build and compile the discriminator\n",
        "        self.discriminator = self.build_discriminator()\n",
        "        self.discriminator.compile(loss='binary_crossentropy',\n",
        "            optimizer=optimizer,\n",
        "            metrics=['accuracy'])\n",
        "\n",
        "        # Build the generator\n",
        "        self.generator = self.build_generator()\n",
        "\n",
        "        # The generator takes noise as input and generates imgs\n",
        "        z = Input(shape=(self.latent_dim,))\n",
        "        img = self.generator(z)\n",
        "\n",
        "        # For the combined model we will only train the generator\n",
        "        self.discriminator.trainable = False\n",
        "\n",
        "        # The discriminator takes generated images as input and determines validity\n",
        "        valid = self.discriminator(img)\n",
        "\n",
        "        # The combined model  (stacked generator and discriminator)\n",
        "        # Trains the generator to fool the discriminator\n",
        "        self.combined = Model(z, valid)\n",
        "        self.combined.compile(loss='binary_crossentropy', optimizer=optimizer)\n",
        "\n",
        "    def build_generator(self):\n",
        "\n",
        "        model = Sequential()\n",
        "\n",
        "        model.add(Dense(128 * 7 * 7, activation=\"relu\", input_dim=self.latent_dim))\n",
        "        model.add(Reshape((7, 7, 128)))\n",
        "        model.add(UpSampling2D())\n",
        "        model.add(Conv2D(128, kernel_size=3, padding=\"same\"))\n",
        "        model.add(BatchNormalization(momentum=0.8))\n",
        "        model.add(Activation(\"relu\"))\n",
        "        model.add(UpSampling2D())\n",
        "        model.add(Conv2D(64, kernel_size=3, padding=\"same\"))\n",
        "        model.add(BatchNormalization(momentum=0.8))\n",
        "        model.add(Activation(\"relu\"))\n",
        "        model.add(Conv2D(self.channels, kernel_size=3, padding=\"same\"))\n",
        "        model.add(Activation(\"tanh\"))\n",
        "\n",
        "        model.summary()\n",
        "\n",
        "        noise = Input(shape=(self.latent_dim,))\n",
        "        img = model(noise)\n",
        "\n",
        "        return Model(noise, img)\n",
        "\n",
        "    def build_discriminator(self):\n",
        "\n",
        "        model = Sequential()\n",
        "\n",
        "        model.add(Conv2D(32, kernel_size=3, strides=2, input_shape=self.img_shape, padding=\"same\"))\n",
        "        model.add(LeakyReLU(alpha=0.2))\n",
        "        model.add(Dropout(0.25))\n",
        "        model.add(Conv2D(64, kernel_size=3, strides=2, padding=\"same\"))\n",
        "        model.add(ZeroPadding2D(padding=((0,1),(0,1))))\n",
        "        model.add(BatchNormalization(momentum=0.8))\n",
        "        model.add(LeakyReLU(alpha=0.2))\n",
        "        model.add(Dropout(0.25))\n",
        "        model.add(Conv2D(128, kernel_size=3, strides=2, padding=\"same\"))\n",
        "        model.add(BatchNormalization(momentum=0.8))\n",
        "        model.add(LeakyReLU(alpha=0.2))\n",
        "        model.add(Dropout(0.25))\n",
        "        model.add(Conv2D(256, kernel_size=3, strides=1, padding=\"same\"))\n",
        "        model.add(BatchNormalization(momentum=0.8))\n",
        "        model.add(LeakyReLU(alpha=0.2))\n",
        "        model.add(Dropout(0.25))\n",
        "        model.add(Flatten())\n",
        "        model.add(Dense(1, activation='sigmoid'))\n",
        "\n",
        "        model.summary()\n",
        "\n",
        "        img = Input(shape=self.img_shape)\n",
        "        validity = model(img)\n",
        "\n",
        "        return Model(img, validity)\n",
        "\n",
        "    def train(self, epochs, batch_size=128, save_interval=50):\n",
        "\n",
        "        # Load the dataset\n",
        "        (X_train, _), (_, _) = mnist.load_data()\n",
        "\n",
        "        # Rescale -1 to 1\n",
        "        X_train = X_train / 127.5 - 1.\n",
        "        X_train = np.expand_dims(X_train, axis=3)\n",
        "\n",
        "        # Adversarial ground truths\n",
        "        valid = np.ones((batch_size, 1))\n",
        "        fake = np.zeros((batch_size, 1))\n",
        "\n",
        "        for epoch in range(epochs):\n",
        "\n",
        "            # ---------------------\n",
        "            #  Train Discriminator\n",
        "            # ---------------------\n",
        "\n",
        "            # Select a random half of images\n",
        "            idx = np.random.randint(0, X_train.shape[0], batch_size)\n",
        "            imgs = X_train[idx]\n",
        "\n",
        "            # Sample noise and generate a batch of new images\n",
        "            noise = np.random.normal(0, 1, (batch_size, self.latent_dim))\n",
        "            gen_imgs = self.generator.predict(noise)\n",
        "\n",
        "            # Train the discriminator (real classified as ones and generated as zeros)\n",
        "            d_loss_real = self.discriminator.train_on_batch(imgs, valid)\n",
        "            d_loss_fake = self.discriminator.train_on_batch(gen_imgs, fake)\n",
        "            d_loss = 0.5 * np.add(d_loss_real, d_loss_fake)\n",
        "\n",
        "            # ---------------------\n",
        "            #  Train Generator\n",
        "            # ---------------------\n",
        "\n",
        "            # Train the generator (wants discriminator to mistake images as real)\n",
        "            g_loss = self.combined.train_on_batch(noise, valid)\n",
        "\n",
        "            # Plot the progress\n",
        "            print (\"%d [D loss: %f, acc.: %.2f%%] [G loss: %f]\" % (epoch, d_loss[0], 100*d_loss[1], g_loss))\n",
        "\n",
        "            # If at save interval => save generated image samples\n",
        "            if epoch % save_interval == 0:\n",
        "                self.save_imgs(epoch)\n",
        "\n",
        "    def save_imgs(self, epoch):\n",
        "        r, c = 5, 5\n",
        "        noise = np.random.normal(0, 1, (r * c, self.latent_dim))\n",
        "        gen_imgs = self.generator.predict(noise)\n",
        "\n",
        "        # Rescale images 0 - 1\n",
        "        gen_imgs = 0.5 * gen_imgs + 0.5\n",
        "\n",
        "        fig, axs = plt.subplots(r, c)\n",
        "        cnt = 0\n",
        "        for i in range(r):\n",
        "            for j in range(c):\n",
        "                axs[i,j].imshow(gen_imgs[cnt, :,:,0], cmap='gray')\n",
        "                axs[i,j].axis('off')\n",
        "                cnt += 1\n",
        "        fig.savefig(\"images/mnist_%d.png\" % epoch)\n",
        "        plt.close()\n",
        "\n",
        "\n",
        "dcgan = DCGAN()\n",
        "dcgan.train(epochs=4000, batch_size=32, save_interval=50)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "BlFy_UjKwsl9",
        "colab_type": "text"
      },
      "source": [
        "## 明星圖片生成"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "-hCQUNgIw1jV",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "def load_dataset(dataset_path, batch_size, image_shape):\n",
        "    dataset_generator = ImageDataGenerator()\n",
        "    dataset_generator = dataset_generator.flow_from_directory(\n",
        "        dataset_path, target_size=(image_shape[0], image_shape[1]),\n",
        "        batch_size=batch_size,\n",
        "        class_mode=None)\n",
        "\n",
        "    return dataset_generator"
      ],
      "execution_count": 11,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "cjUQhuviw24Y",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "def build_generator():\n",
        "\n",
        "    model = Sequential()\n",
        "\n",
        "    model.add(Dense(units=4 * 4 * 512,\n",
        "                   input_shape=(1, 1, 100)))\n",
        "    model.add(Reshape(target_shape=(4, 4, 512)))\n",
        "    model.add(BatchNormalization(momentum=0.5))\n",
        "    model.add(Activation('relu'))\n",
        "\n",
        "    model.add(Conv2DTranspose(filters=256, kernel_size=(5, 5),\n",
        "                                  strides=(2, 2), padding='same'))\n",
        "    model.add(BatchNormalization(momentum=0.5))\n",
        "    model.add(Activation('relu'))\n",
        "\n",
        "    model.add(Conv2DTranspose(filters=128, kernel_size=(5, 5),\n",
        "                                  strides=(2, 2), padding='same'))\n",
        "    model.add(BatchNormalization(momentum=0.5))\n",
        "    model.add(Activation('relu'))\n",
        "\n",
        "    model.add(Conv2DTranspose(filters=64, kernel_size=(5, 5),\n",
        "                                  strides=(2, 2), padding='same'))\n",
        "    model.add(BatchNormalization(momentum=0.5))\n",
        "    model.add(Activation('relu'))\n",
        "\n",
        "    model.add(Conv2DTranspose(filters=3, kernel_size=(5, 5),\n",
        "                                  strides=(2, 2), padding='same'))\n",
        "    model.add(Activation('tanh'))\n",
        "    model.summary()\n",
        "    optimizer = Adam(lr=0.00015, beta_1=0.5)\n",
        "    model.compile(loss='binary_crossentropy',\n",
        "                      optimizer=optimizer,\n",
        "                      metrics=None)\n",
        "\n",
        "    return model"
      ],
      "execution_count": 12,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "U3eTvYIuw-N4",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "def build_discriminator():\n",
        "\n",
        "    model = Sequential()\n",
        "    \n",
        "    image_shape = (64, 64, 3)\n",
        "    model.add(Conv2D(filters=64, kernel_size=(5, 5),\n",
        "                             strides=(2, 2), padding='same',\n",
        "                             input_shape=(image_shape)))\n",
        "    model.add(LeakyReLU(0.2))\n",
        "\n",
        "    model.add(Conv2D(filters=128, kernel_size=(5, 5),\n",
        "                             strides=(2, 2), padding='same'))\n",
        "    model.add(BatchNormalization(momentum=0.5))\n",
        "    model.add(LeakyReLU(0.2))\n",
        "\n",
        "    model.add(Conv2D(filters=256, kernel_size=(5, 5),\n",
        "                             strides=(2, 2), padding='same'))\n",
        "    model.add(BatchNormalization(momentum=0.5))\n",
        "    model.add(LeakyReLU(0.2))\n",
        "\n",
        "    model.add(Conv2D(filters=512, kernel_size=(5, 5),\n",
        "                             strides=(2, 2), padding='same'))\n",
        "    model.add(BatchNormalization(momentum=0.5))\n",
        "    model.add(LeakyReLU(0.2))\n",
        "\n",
        "    model.add(Flatten())\n",
        "    model.add(Dense(1))\n",
        "    model.add(Activation('sigmoid'))\n",
        "    model.summary()\n",
        "\n",
        "    optimizer = Adam(lr=0.0002, beta_1=0.5)\n",
        "    model.compile(loss='binary_crossentropy',\n",
        "                          optimizer=optimizer,\n",
        "                          metrics=None)\n",
        "\n",
        "    return model"
      ],
      "execution_count": 13,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "LU2uqOphxI13",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "def sample_images(generated_images, epoch):\n",
        "\n",
        "    plt.figure(figsize=(8, 8), num=2)\n",
        "    gs1 = gridspec.GridSpec(8, 8)\n",
        "    gs1.update(wspace=0, hspace=0)\n",
        "\n",
        "    for i in range(generated_images.shape[0]):\n",
        "        ax1 = plt.subplot(gs1[i])\n",
        "        ax1.set_aspect('equal')\n",
        "        image = generated_images[i, :, :, :]\n",
        "        image += 1\n",
        "        image *= 127.5\n",
        "        fig = plt.imshow(image.astype(np.uint8))\n",
        "        plt.axis('off')\n",
        "        fig.axes.get_xaxis().set_visible(False)\n",
        "        fig.axes.get_yaxis().set_visible(False)\n",
        "\n",
        "    plt.tight_layout()\n",
        "    save_name = 'images/generatedSamples_epoch' + str(\n",
        "        epoch + 1) + '.png'\n",
        "\n",
        "    plt.savefig(save_name, bbox_inches='tight', pad_inches=0)\n",
        "    plt.pause(0.0000000001)\n",
        "    plt.show()"
      ],
      "execution_count": 14,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "POpBK5rwxTDY",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 676
        },
        "outputId": "6e9447b1-e376-492d-a7e5-f6e3ac5818ee"
      },
      "source": [
        "from keras.layers import Conv2D, Conv2DTranspose, Reshape\n",
        "generator = build_generator()"
      ],
      "execution_count": 18,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Model: \"sequential_2\"\n",
            "_________________________________________________________________\n",
            "Layer (type)                 Output Shape              Param #   \n",
            "=================================================================\n",
            "dense_1 (Dense)              (None, 1, 1, 8192)        827392    \n",
            "_________________________________________________________________\n",
            "reshape_1 (Reshape)          (None, 4, 4, 512)         0         \n",
            "_________________________________________________________________\n",
            "batch_normalization_4 (Batch (None, 4, 4, 512)         2048      \n",
            "_________________________________________________________________\n",
            "activation_5 (Activation)    (None, 4, 4, 512)         0         \n",
            "_________________________________________________________________\n",
            "conv2d_transpose_4 (Conv2DTr (None, 8, 8, 256)         3277056   \n",
            "_________________________________________________________________\n",
            "batch_normalization_5 (Batch (None, 8, 8, 256)         1024      \n",
            "_________________________________________________________________\n",
            "activation_6 (Activation)    (None, 8, 8, 256)         0         \n",
            "_________________________________________________________________\n",
            "conv2d_transpose_5 (Conv2DTr (None, 16, 16, 128)       819328    \n",
            "_________________________________________________________________\n",
            "batch_normalization_6 (Batch (None, 16, 16, 128)       512       \n",
            "_________________________________________________________________\n",
            "activation_7 (Activation)    (None, 16, 16, 128)       0         \n",
            "_________________________________________________________________\n",
            "conv2d_transpose_6 (Conv2DTr (None, 32, 32, 64)        204864    \n",
            "_________________________________________________________________\n",
            "batch_normalization_7 (Batch (None, 32, 32, 64)        256       \n",
            "_________________________________________________________________\n",
            "activation_8 (Activation)    (None, 32, 32, 64)        0         \n",
            "_________________________________________________________________\n",
            "conv2d_transpose_7 (Conv2DTr (None, 64, 64, 3)         4803      \n",
            "_________________________________________________________________\n",
            "activation_9 (Activation)    (None, 64, 64, 3)         0         \n",
            "=================================================================\n",
            "Total params: 5,137,283\n",
            "Trainable params: 5,135,363\n",
            "Non-trainable params: 1,920\n",
            "_________________________________________________________________\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "cabn55CvxVn1",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 641
        },
        "outputId": "4cf78c22-7de3-456c-9c82-bee5f9b386db"
      },
      "source": [
        "discriminator = build_discriminator()"
      ],
      "execution_count": 19,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Model: \"sequential_3\"\n",
            "_________________________________________________________________\n",
            "Layer (type)                 Output Shape              Param #   \n",
            "=================================================================\n",
            "conv2d_1 (Conv2D)            (None, 32, 32, 64)        4864      \n",
            "_________________________________________________________________\n",
            "leaky_re_lu (LeakyReLU)      (None, 32, 32, 64)        0         \n",
            "_________________________________________________________________\n",
            "conv2d_2 (Conv2D)            (None, 16, 16, 128)       204928    \n",
            "_________________________________________________________________\n",
            "batch_normalization_8 (Batch (None, 16, 16, 128)       512       \n",
            "_________________________________________________________________\n",
            "leaky_re_lu_1 (LeakyReLU)    (None, 16, 16, 128)       0         \n",
            "_________________________________________________________________\n",
            "conv2d_3 (Conv2D)            (None, 8, 8, 256)         819456    \n",
            "_________________________________________________________________\n",
            "batch_normalization_9 (Batch (None, 8, 8, 256)         1024      \n",
            "_________________________________________________________________\n",
            "leaky_re_lu_2 (LeakyReLU)    (None, 8, 8, 256)         0         \n",
            "_________________________________________________________________\n",
            "conv2d_4 (Conv2D)            (None, 4, 4, 512)         3277312   \n",
            "_________________________________________________________________\n",
            "batch_normalization_10 (Batc (None, 4, 4, 512)         2048      \n",
            "_________________________________________________________________\n",
            "leaky_re_lu_3 (LeakyReLU)    (None, 4, 4, 512)         0         \n",
            "_________________________________________________________________\n",
            "flatten (Flatten)            (None, 8192)              0         \n",
            "_________________________________________________________________\n",
            "dense_2 (Dense)              (None, 1)                 8193      \n",
            "_________________________________________________________________\n",
            "activation_10 (Activation)   (None, 1)                 0         \n",
            "=================================================================\n",
            "Total params: 4,318,337\n",
            "Trainable params: 4,316,545\n",
            "Non-trainable params: 1,792\n",
            "_________________________________________________________________\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "GDLFYbBn1BVu",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "! unzip celeba.zip"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "FMbO-QmW1oK1",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "! mkdir celeba/train"
      ],
      "execution_count": 23,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "B0RoS8QR1sG0",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "! mv celeba/*.jpg celeba/train/"
      ],
      "execution_count": 24,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "yC7mbo9bxhmF",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 35
        },
        "outputId": "b1f8d891-125d-415a-d586-4c2fbde150a5"
      },
      "source": [
        "dataset_path = 'celeba'\n",
        "batch_size   =  64\n",
        "image_shape  = (64, 64, 3)\n",
        "from keras.preprocessing.image import ImageDataGenerator\n",
        "dataset_generator = load_dataset(dataset_path, batch_size, image_shape)"
      ],
      "execution_count": 25,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Found 9521 images belonging to 1 classes.\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "RfcK7rRG1FYW",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "\n",
        "number_of_batches  = int(dataset_generator.samples / batch_size)\n",
        "adversarial_loss   = np.empty(shape=1)\n",
        "discriminator_loss = np.empty(shape=1)\n",
        "epochs = 250"
      ],
      "execution_count": 26,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "eTeg2QWe11O0",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "gan = Sequential()\n",
        "\n",
        "discriminator.trainable = False\n",
        "gan.add(generator)\n",
        "gan.add(discriminator)\n",
        "\n",
        "optimizer = Adam(lr=0.00015, beta_1=0.5)\n",
        "gan.compile(loss='binary_crossentropy', optimizer=optimizer,\n",
        "                    metrics=None)"
      ],
      "execution_count": 27,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "RqqSH9jG13pU",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "import os\n",
        "if not os.path.exists('models'):\n",
        "    os.mkdir('models')"
      ],
      "execution_count": 28,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "n1UT7Khq157z",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "import time\n",
        "import os\n",
        "import matplotlib.pyplot as plt\n",
        "import matplotlib.gridspec as gridspec\n",
        "import numpy as np\n",
        "\n",
        "for epoch in range(epochs):\n",
        "    print(\"Epoch {} / {}:\".format(epoch+1, epochs))\n",
        "\n",
        "    for batch_number in range(number_of_batches):\n",
        "        # 取得一批图片数据\n",
        "        real_images = dataset_generator.next()\n",
        "        # 将图片标准化至-1到1\n",
        "        real_images = real_images / 127.5 - 1.\n",
        "\n",
        "        # 每批大小不同，需每次计算该批数据量\n",
        "        current_batch_size = real_images.shape[0]\n",
        "\n",
        "        # 产生噪音\n",
        "        noise = np.random.normal(0, 1, size=(current_batch_size,) + (1, 1, 100))\n",
        "\n",
        "        # 使用生成器生成图片\n",
        "        generated_images = generator.predict(noise)\n",
        "\n",
        "        # 增加噪音扰乱鉴别器\n",
        "        real_y = (np.ones(current_batch_size) - np.random.random_sample(current_batch_size) * 0.2)\n",
        "        fake_y = np.random.random_sample(current_batch_size) * 0.2\n",
        "\n",
        "        # 训练鉴别器\n",
        "        discriminator.trainable = True\n",
        "\n",
        "        d_loss = discriminator.train_on_batch(real_images, real_y)\n",
        "        d_loss += discriminator.train_on_batch(generated_images, fake_y)\n",
        "\n",
        "        discriminator_loss = np.append(discriminator_loss, d_loss)\n",
        "\n",
        "        # 训练生成器\n",
        "        discriminator.trainable = False\n",
        "\n",
        "        noise = np.random.normal(0, 1,size=(current_batch_size * 2,) +(1, 1, 100))\n",
        "\n",
        "        # 给予反向标签以误导鉴别器\n",
        "        fake_y = (np.ones(current_batch_size * 2) -\n",
        "                      np.random.random_sample(current_batch_size * 2) * 0.2)\n",
        "\n",
        "        g_loss = gan.train_on_batch(noise, fake_y)\n",
        "        adversarial_loss = np.append(adversarial_loss, g_loss)\n",
        "    \n",
        "\n",
        "        # 打印损失\n",
        "        print(\" generator loss {} | discriminator loss {} \".format(g_loss, d_loss))\n",
        "\n",
        "\n",
        "    # 存储模型与生成图片\n",
        "    if (epoch + 1) % 3 == 0:\n",
        "        discriminator.trainable = True\n",
        "        generator.save('models/generator_epoch' + str(epoch) + '.hdf5')\n",
        "        discriminator.save('models/discriminator_epoch' +str(epoch) + '.hdf5')\n",
        "        \n",
        "        sample_images(generated_images, epoch)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "zbiRje-f4TaK",
        "colab_type": "text"
      },
      "source": [
        "## Cycle GAN"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "ZBEKR6sO_CB3",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 332
        },
        "outputId": "5421fadb-6d58-4218-cdc8-21c3b3af1f77"
      },
      "source": [
        "! pip install scipy==1.2.1"
      ],
      "execution_count": 139,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Collecting scipy==1.2.1\n",
            "\u001b[?25l  Downloading https://files.pythonhosted.org/packages/7f/5f/c48860704092933bf1c4c1574a8de1ffd16bf4fde8bab190d747598844b2/scipy-1.2.1-cp36-cp36m-manylinux1_x86_64.whl (24.8MB)\n",
            "\u001b[K     |████████████████████████████████| 24.8MB 130kB/s \n",
            "\u001b[?25hRequirement already satisfied: numpy>=1.8.2 in /usr/local/lib/python3.6/dist-packages (from scipy==1.2.1) (1.18.5)\n",
            "\u001b[31mERROR: umap-learn 0.4.6 has requirement scipy>=1.3.1, but you'll have scipy 1.2.1 which is incompatible.\u001b[0m\n",
            "\u001b[31mERROR: tensorflow 2.3.0 has requirement scipy==1.4.1, but you'll have scipy 1.2.1 which is incompatible.\u001b[0m\n",
            "\u001b[31mERROR: albumentations 0.1.12 has requirement imgaug<0.2.7,>=0.2.5, but you'll have imgaug 0.2.9 which is incompatible.\u001b[0m\n",
            "Installing collected packages: scipy\n",
            "  Found existing installation: scipy 1.4.1\n",
            "    Uninstalling scipy-1.4.1:\n",
            "      Successfully uninstalled scipy-1.4.1\n",
            "Successfully installed scipy-1.2.1\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "display_data",
          "data": {
            "application/vnd.colab-display-data+json": {
              "pip_warning": {
                "packages": [
                  "scipy"
                ]
              }
            }
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "jlsQSL115Bns",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 35
        },
        "outputId": "acdd6b49-c057-40ac-8b50-e0f1b5dacaeb"
      },
      "source": [
        "% tensorflow_version 1.x"
      ],
      "execution_count": 44,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "TensorFlow is already loaded. Please restart the runtime to change versions.\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "saQzLdfH2Ovy",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 315
        },
        "outputId": "7010c32a-1986-4c38-b3b1-8fb923316f2d"
      },
      "source": [
        "! pip install git+https://www.github.com/keras-team/keras-contrib.git"
      ],
      "execution_count": 45,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Collecting git+https://www.github.com/keras-team/keras-contrib.git\n",
            "  Cloning https://www.github.com/keras-team/keras-contrib.git to /tmp/pip-req-build-wizg8j3e\n",
            "  Running command git clone -q https://www.github.com/keras-team/keras-contrib.git /tmp/pip-req-build-wizg8j3e\n",
            "Requirement already satisfied: keras in /usr/local/lib/python3.6/dist-packages (from keras-contrib==2.0.8) (2.4.3)\n",
            "Requirement already satisfied: numpy>=1.9.1 in /usr/local/lib/python3.6/dist-packages (from keras->keras-contrib==2.0.8) (1.18.5)\n",
            "Requirement already satisfied: h5py in /usr/local/lib/python3.6/dist-packages (from keras->keras-contrib==2.0.8) (2.10.0)\n",
            "Requirement already satisfied: pyyaml in /usr/local/lib/python3.6/dist-packages (from keras->keras-contrib==2.0.8) (3.13)\n",
            "Requirement already satisfied: scipy>=0.14 in /usr/local/lib/python3.6/dist-packages (from keras->keras-contrib==2.0.8) (1.4.1)\n",
            "Requirement already satisfied: six in /usr/local/lib/python3.6/dist-packages (from h5py->keras->keras-contrib==2.0.8) (1.15.0)\n",
            "Building wheels for collected packages: keras-contrib\n",
            "  Building wheel for keras-contrib (setup.py) ... \u001b[?25l\u001b[?25hdone\n",
            "  Created wheel for keras-contrib: filename=keras_contrib-2.0.8-cp36-none-any.whl size=101064 sha256=a102f11ed3d41a3237d9474bff5f20eb7ffdd75d065d8e2d6d5743e178bfe759\n",
            "  Stored in directory: /tmp/pip-ephem-wheel-cache-581kfzln/wheels/11/27/c8/4ed56de7b55f4f61244e2dc6ef3cdbaff2692527a2ce6502ba\n",
            "Successfully built keras-contrib\n",
            "Installing collected packages: keras-contrib\n",
            "Successfully installed keras-contrib-2.0.8\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "T_LALKjZ4tpo",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "from keras_contrib.layers.normalization import instancenormalization"
      ],
      "execution_count": 46,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "HJ9t5MMf6EIR",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "import scipy\n",
        "\n",
        "from keras.datasets import mnist\n",
        "from keras_contrib.layers.normalization.instancenormalization import InstanceNormalization\n",
        "from keras.layers import Input, Dense, Reshape, Flatten, Dropout, Concatenate\n",
        "from keras.layers import BatchNormalization, Activation, ZeroPadding2D\n",
        "from keras.layers.advanced_activations import LeakyReLU\n",
        "from keras.layers.convolutional import UpSampling2D, Conv2D\n",
        "from keras.models import Sequential, Model\n",
        "from keras.optimizers import Adam\n",
        "import datetime\n",
        "import matplotlib.pyplot as plt\n",
        "import sys\n",
        "import numpy as np\n",
        "import os"
      ],
      "execution_count": 47,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "c56JqRic6GIC",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "from glob import glob\n",
        "import numpy as np\n",
        "import scipy.misc\n",
        "from skimage import io\n",
        "\n",
        "class DataLoader():\n",
        "    def __init__(self, dataset_name, img_res=(128, 128)):\n",
        "        self.dataset_name = dataset_name\n",
        "        self.img_res = img_res\n",
        "\n",
        "    def load_data(self, domain, batch_size=1, is_testing=False):\n",
        "        data_type = \"train%s\" % domain if not is_testing else \"test%s\" % domain\n",
        "        path = glob('./datasets/%s/%s/*' % (self.dataset_name, data_type))\n",
        "        #print(path)\n",
        "        batch_images = np.random.choice(path, size=batch_size)\n",
        "\n",
        "        imgs = []\n",
        "        for img_path in batch_images:\n",
        "            img = self.imread(img_path)\n",
        "            if not is_testing:\n",
        "                img = scipy.misc.imresize(img, self.img_res)\n",
        "\n",
        "                if np.random.random() > 0.5:\n",
        "                    img = np.fliplr(img)\n",
        "            else:\n",
        "                img = scipy.misc.imresize(img, self.img_res)\n",
        "            imgs.append(img)\n",
        "\n",
        "        imgs = np.array(imgs)/127.5 - 1.\n",
        "\n",
        "        return imgs\n",
        "\n",
        "    def load_batch(self, batch_size=1, is_testing=False):\n",
        "        data_type = \"train\" if not is_testing else \"val\"\n",
        "        path_A = glob('./datasets/%s/%sA/*' % (self.dataset_name, data_type))\n",
        "        path_B = glob('./datasets/%s/%sB/*' % (self.dataset_name, data_type))\n",
        "        #print(path_A)\n",
        "        self.n_batches = int(min(len(path_A), len(path_B)) / batch_size)\n",
        "        total_samples = self.n_batches * batch_size\n",
        "\n",
        "        # Sample n_batches * batch_size from each path list so that model sees all\n",
        "        # samples from both domains\n",
        "        path_A = np.random.choice(path_A, total_samples, replace=False)\n",
        "        path_B = np.random.choice(path_B, total_samples, replace=False)\n",
        "\n",
        "        for i in range(self.n_batches-1):\n",
        "            batch_A = path_A[i*batch_size:(i+1)*batch_size]\n",
        "            batch_B = path_B[i*batch_size:(i+1)*batch_size]\n",
        "            imgs_A, imgs_B = [], []\n",
        "            for img_A, img_B in zip(batch_A, batch_B):\n",
        "                img_A = self.imread(img_A)\n",
        "                img_B = self.imread(img_B)\n",
        "\n",
        "                img_A = scipy.misc.imresize(img_A, self.img_res)\n",
        "                img_B = scipy.misc.imresize(img_B, self.img_res)\n",
        "\n",
        "                if not is_testing and np.random.random() > 0.5:\n",
        "                        img_A = np.fliplr(img_A)\n",
        "                        img_B = np.fliplr(img_B)\n",
        "\n",
        "                imgs_A.append(img_A)\n",
        "                imgs_B.append(img_B)\n",
        "\n",
        "            imgs_A = np.array(imgs_A)/127.5 - 1.\n",
        "            imgs_B = np.array(imgs_B)/127.5 - 1.\n",
        "\n",
        "            yield imgs_A, imgs_B\n",
        "\n",
        "    def load_img(self, path):\n",
        "        img = self.imread(path)\n",
        "        img = scipy.misc.imresize(img, self.img_res)\n",
        "        img = img/127.5 - 1.\n",
        "        return img[np.newaxis, :, :, :]\n",
        "\n",
        "    def imread(self, path):\n",
        "        return scipy.misc.imread(path, mode='RGB').astype(np.float)"
      ],
      "execution_count": 140,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "HqiwWbl66peI",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 176
        },
        "outputId": "978f14f3-c39e-4a82-a8c6-caa8361e01ba"
      },
      "source": [
        "! wget https://people.eecs.berkeley.edu/~taesung_park/CycleGAN/datasets/monet2photo.zip"
      ],
      "execution_count": 117,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "--2020-08-22 07:01:57--  https://people.eecs.berkeley.edu/~taesung_park/CycleGAN/datasets/monet2photo.zip\n",
            "Resolving people.eecs.berkeley.edu (people.eecs.berkeley.edu)... 128.32.189.73\n",
            "Connecting to people.eecs.berkeley.edu (people.eecs.berkeley.edu)|128.32.189.73|:443... connected.\n",
            "HTTP request sent, awaiting response... 200 OK\n",
            "Length: 305231073 (291M) [application/zip]\n",
            "Saving to: ‘monet2photo.zip.1’\n",
            "\n",
            "monet2photo.zip.1     0%[                    ]  40.00K   131KB/s               ^C\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "gtRkFJOw6xWU",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "! unzip monet2photo.zip"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Sk-Gdf9U9EAd",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "! mv monet2photo datasets"
      ],
      "execution_count": 91,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "3mwf6tV59Yuc",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "! mkdir datasets/monet2photo\n",
        "! mv datasets/trainA datasets/monet2photo\n",
        "! mv datasets/trainB datasets/monet2photo\n",
        "! mv datasets/testA datasets/monet2photo\n",
        "! mv datasets/testB datasets/monet2photo"
      ],
      "execution_count": 102,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "KgAEn65o8tEB",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "a = DataLoader('monet2photo')"
      ],
      "execution_count": 129,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "qx9oO9sN6Yk_",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "# 图片参数\n",
        "img_rows = 128\n",
        "img_cols = 128\n",
        "channels = 3\n",
        "img_shape = (img_rows, img_cols, channels)\n",
        "dataset_name = 'monet2photo'\n",
        "data_loader = DataLoader(dataset_name=dataset_name,\n",
        "             img_res=(img_rows, img_cols))"
      ],
      "execution_count": 135,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "hoQdAyXO6Z1w",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "def build_generator():\n",
        "    \"\"\"U-Net Generator\"\"\"\n",
        "\n",
        "    def conv2d(layer_input, filters, f_size=4):\n",
        "        \"\"\"Layers used during downsampling\"\"\"\n",
        "        d = Conv2D(filters, kernel_size=f_size, strides=2, padding='same')(layer_input)\n",
        "        d = LeakyReLU(alpha=0.2)(d)\n",
        "        d = InstanceNormalization()(d)\n",
        "        return d\n",
        "\n",
        "    def deconv2d(layer_input, skip_input, filters, f_size=4, dropout_rate=0):\n",
        "        \"\"\"Layers used during upsampling\"\"\"\n",
        "        u = UpSampling2D(size=2)(layer_input)\n",
        "        u = Conv2D(filters, kernel_size=f_size, strides=1, padding='same', activation='relu')(u)\n",
        "        if dropout_rate:\n",
        "            u = Dropout(dropout_rate)(u)\n",
        "        u = InstanceNormalization()(u)\n",
        "        u = Concatenate()([u, skip_input])\n",
        "        return u\n",
        "\n",
        "    # Image input\n",
        "    d0 = Input(shape=img_shape)\n",
        "\n",
        "    # Downsampling\n",
        "    d1 = conv2d(d0, gf)\n",
        "    d2 = conv2d(d1, gf*2)\n",
        "    d3 = conv2d(d2, gf*4)\n",
        "    d4 = conv2d(d3, gf*8)\n",
        "\n",
        "    # Upsampling\n",
        "    u1 = deconv2d(d4, d3, gf*4)\n",
        "    u2 = deconv2d(u1, d2, gf*2)\n",
        "    u3 = deconv2d(u2, d1, gf)\n",
        "\n",
        "    u4 = UpSampling2D(size=2)(u3)\n",
        "    output_img = Conv2D(channels, kernel_size=4, strides=1, padding='same', activation='tanh')(u4)\n",
        "\n",
        "    return Model(d0, output_img)"
      ],
      "execution_count": 136,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "YpISrqe47KC_",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "from keras.layers import Input, Dense, Reshape, Flatten, Dropout\n",
        "def build_discriminator():\n",
        "\n",
        "    def d_layer(layer_input, filters, f_size=4, normalization=True):\n",
        "        \"\"\"Discriminator layer\"\"\"\n",
        "        d = Conv2D(filters, kernel_size=f_size, strides=2, padding='same')(layer_input)\n",
        "        d = LeakyReLU(alpha=0.2)(d)\n",
        "        if normalization:\n",
        "            d = InstanceNormalization()(d)\n",
        "        return d\n",
        "\n",
        "    img = Input(shape=img_shape)\n",
        "\n",
        "    d1 = d_layer(img, df, normalization=False)\n",
        "    d2 = d_layer(d1, df*2)\n",
        "    d3 = d_layer(d2, df*4)\n",
        "    d4 = d_layer(d3, df*8)\n",
        "\n",
        "    validity = Conv2D(1, kernel_size=4, strides=1, padding='same')(d4)\n",
        "    return Model(img, validity)"
      ],
      "execution_count": 137,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "8Xm4poID70Pr",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "def train(epochs, batch_size=1, sample_interval=50):\n",
        "\n",
        "    start_time = datetime.datetime.now()\n",
        "\n",
        "    # Adversarial loss ground truths\n",
        "    valid = np.ones((batch_size,) + disc_patch)\n",
        "    fake = np.zeros((batch_size,) + disc_patch)\n",
        "\n",
        "    for epoch in range(epochs):\n",
        "        print(epoch)\n",
        "        for batch_i, (imgs_A, imgs_B) in enumerate(data_loader.load_batch(batch_size)):\n",
        "\n",
        "            # ----------------------\n",
        "            #  Train Discriminators\n",
        "            # ----------------------\n",
        "\n",
        "            # Translate images to opposite domain\n",
        "            fake_B = g_AB.predict(imgs_A)\n",
        "            fake_A = g_BA.predict(imgs_B)\n",
        "\n",
        "            # Train the discriminators (original images = real / translated = Fake)\n",
        "            dA_loss_real = d_A.train_on_batch(imgs_A, valid)\n",
        "            dA_loss_fake = d_A.train_on_batch(fake_A, fake)\n",
        "            dA_loss = 0.5 * np.add(dA_loss_real, dA_loss_fake)\n",
        "\n",
        "            dB_loss_real = d_B.train_on_batch(imgs_B, valid)\n",
        "            dB_loss_fake = d_B.train_on_batch(fake_B, fake)\n",
        "            dB_loss = 0.5 * np.add(dB_loss_real, dB_loss_fake)\n",
        "\n",
        "            # Total disciminator loss\n",
        "            d_loss = 0.5 * np.add(dA_loss, dB_loss)\n",
        "\n",
        "\n",
        "            # ------------------\n",
        "            #  Train Generators\n",
        "            # ------------------\n",
        "\n",
        "            # Train the generators\n",
        "            g_loss = combined.train_on_batch([imgs_A, imgs_B],\n",
        "                              [valid, valid,\n",
        "                              imgs_A, imgs_B,\n",
        "                              imgs_A, imgs_B])\n",
        "\n",
        "            elapsed_time = datetime.datetime.now() - start_time\n",
        "\n",
        "            # Plot the progress\n",
        "            print (\"[Epoch %d/%d] [Batch %d/%d] [D loss: %f, acc: %3d%%] [G loss: %05f, adv: %05f, recon: %05f, id: %05f] time: %s \" \\\n",
        "                                                                    % ( epoch, epochs,\n",
        "                                                                        batch_i, data_loader.n_batches,\n",
        "                                                                        d_loss[0], 100*d_loss[1],\n",
        "                                                                        g_loss[0],\n",
        "                                                                        np.mean(g_loss[1:3]),\n",
        "                                                                        np.mean(g_loss[3:5]),\n",
        "                                                                        np.mean(g_loss[5:6]),\n",
        "                                                                        elapsed_time))\n",
        "\n",
        "            # If at save interval => save generated image samples\n",
        "            if batch_i % sample_interval == 0:\n",
        "                sample_images(epoch, batch_i)"
      ],
      "execution_count": 108,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "S_7wN_SQ8L-5",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "def sample_images(epoch, batch_i):\n",
        "    os.makedirs('images/%s' % dataset_name, exist_ok=True)\n",
        "    r, c = 2, 3\n",
        "\n",
        "    imgs_A = data_loader.load_data(domain=\"A\", batch_size=1, is_testing=True)\n",
        "    imgs_B = data_loader.load_data(domain=\"B\", batch_size=1, is_testing=True)\n",
        "    print(img_A)\n",
        "    # Demo (for GIF)\n",
        "    #imgs_A = data_loader.load_img('datasets/apple2orange/testA/n07740461_1541.jpg')\n",
        "    #imgs_B = data_loader.load_img('datasets/apple2orange/testB/n07749192_4241.jpg')\n",
        "\n",
        "    # Translate images to the other domain\n",
        "    fake_B = g_AB.predict(imgs_A)\n",
        "    fake_A = g_BA.predict(imgs_B)\n",
        "    # Translate back to original domain\n",
        "    reconstr_A = g_BA.predict(fake_B)\n",
        "    reconstr_B = g_AB.predict(fake_A)\n",
        "\n",
        "    gen_imgs = np.concatenate([imgs_A, fake_B, reconstr_A, imgs_B, fake_A, reconstr_B])\n",
        "\n",
        "    # Rescale images 0 - 1\n",
        "    gen_imgs = 0.5 * gen_imgs + 0.5\n",
        "\n",
        "    titles = ['Original', 'Translated', 'Reconstructed']\n",
        "    fig, axs = plt.subplots(r, c)\n",
        "    cnt = 0\n",
        "    for i in range(r):\n",
        "        for j in range(c):\n",
        "            axs[i,j].imshow(gen_imgs[cnt])\n",
        "            axs[i, j].set_title(titles[j])\n",
        "            axs[i,j].axis('off')\n",
        "            cnt += 1\n",
        "    fig.savefig(\"images/%s/%d_%d.png\" % (dataset_name, epoch, batch_i))\n",
        "    plt.close()"
      ],
      "execution_count": 98,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "yp_q9yPp8Oux",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "# Calculate output shape of D (PatchGAN)\n",
        "patch = int(img_rows / 2**4)\n",
        "disc_patch = (patch, patch, 1)\n",
        "\n",
        "# Number of filters in the first layer of G and D\n",
        "gf = 32\n",
        "df = 64\n",
        "\n",
        "# Loss weights\n",
        "lambda_cycle = 10.0                    # Cycle-consistency loss\n",
        "lambda_id = 0.1 * lambda_cycle    # Identity loss\n",
        "\n",
        "optimizer = Adam(0.0002, 0.5)\n",
        "\n",
        "# Build and compile the discriminators\n",
        "d_A = build_discriminator()\n",
        "d_B = build_discriminator()\n",
        "d_A.compile(loss='mse',\n",
        "    optimizer=optimizer,\n",
        "    metrics=['accuracy'])\n",
        "d_B.compile(loss='mse',\n",
        "    optimizer=optimizer,\n",
        "    metrics=['accuracy'])"
      ],
      "execution_count": 99,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "yTgLF8HU8Sjp",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "#-------------------------\n",
        "# Construct Computational\n",
        "#   Graph of Generators\n",
        "#-------------------------\n",
        "\n",
        "# Build the generators\n",
        "g_AB = build_generator()\n",
        "g_BA = build_generator()\n",
        "\n",
        "# Input images from both domains\n",
        "img_A = Input(shape=img_shape)\n",
        "img_B = Input(shape=img_shape)\n",
        "\n",
        "# Translate images to the other domain\n",
        "fake_B = g_AB(img_A)\n",
        "fake_A = g_BA(img_B)\n",
        "# Translate images back to original domain\n",
        "reconstr_A = g_BA(fake_B)\n",
        "reconstr_B = g_AB(fake_A)\n",
        "# Identity mapping of images\n",
        "img_A_id = g_BA(img_A)\n",
        "img_B_id = g_AB(img_B)\n",
        "\n",
        "# For the combined model we will only train the generators\n",
        "d_A.trainable = False\n",
        "d_B.trainable = False\n",
        "\n",
        "# Discriminators determines validity of translated images\n",
        "valid_A = d_A(fake_A)\n",
        "valid_B = d_B(fake_B)\n",
        "\n",
        "# Combined model trains generators to fool discriminators\n",
        "combined = Model(inputs=[img_A, img_B],\n",
        "                      outputs=[ valid_A, valid_B,\n",
        "                                reconstr_A, reconstr_B,\n",
        "                                img_A_id, img_B_id ])\n",
        "combined.compile(loss=['mse', 'mse',\n",
        "                       'mae', 'mae',\n",
        "                       'mae', 'mae'],\n",
        "                    loss_weights=[  1, 1,\n",
        "                                    lambda_cycle, lambda_cycle,\n",
        "                                    lambda_id, lambda_id ],\n",
        "                    optimizer=optimizer)"
      ],
      "execution_count": 100,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "44XVgrda8WT4",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 364
        },
        "outputId": "5bd34c79-4bfe-41e7-90a3-b6ad3eecae37"
      },
      "source": [
        "train(epochs=20, batch_size=10, sample_interval=10)"
      ],
      "execution_count": 101,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "0\n",
            "1\n",
            "2\n",
            "3\n",
            "4\n",
            "5\n",
            "6\n",
            "7\n",
            "8\n",
            "9\n",
            "10\n",
            "11\n",
            "12\n",
            "13\n",
            "14\n",
            "15\n",
            "16\n",
            "17\n",
            "18\n",
            "19\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "HFhPKn4-_uUW",
        "colab_type": "text"
      },
      "source": [
        "## Pretained Model"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Rn8aJ9-38YiI",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 121
        },
        "outputId": "0fc7cf7b-275e-445b-dbed-04d1057fc02f"
      },
      "source": [
        "%cd /content/\n",
        "! git clone https://github.com/junyanz/pytorch-CycleGAN-and-pix2pix.git"
      ],
      "execution_count": 154,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "/content\n",
            "Cloning into 'pytorch-CycleGAN-and-pix2pix'...\n",
            "remote: Enumerating objects: 2318, done.\u001b[K\n",
            "remote: Total 2318 (delta 0), reused 0 (delta 0), pack-reused 2318\u001b[K\n",
            "Receiving objects: 100% (2318/2318), 8.08 MiB | 6.52 MiB/s, done.\n",
            "Resolving deltas: 100% (1490/1490), done.\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "73UtZzbv_wF9",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 332
        },
        "outputId": "89debdb3-fd2e-41e5-fd16-ae363f0b704a"
      },
      "source": [
        "%cd /content/pytorch-CycleGAN-and-pix2pix\n",
        "! bash ./scripts/download_cyclegan_model.sh style_cezanne"
      ],
      "execution_count": 156,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "/content/pytorch-CycleGAN-and-pix2pix\n",
            "Note: available models are apple2orange, orange2apple, summer2winter_yosemite, winter2summer_yosemite, horse2zebra, zebra2horse, monet2photo, style_monet, style_cezanne, style_ukiyoe, style_vangogh, sat2map, map2sat, cityscapes_photo2label, cityscapes_label2photo, facades_photo2label, facades_label2photo, iphone2dslr_flower\n",
            "Specified [style_cezanne]\n",
            "WARNING: timestamping does nothing in combination with -O. See the manual\n",
            "for details.\n",
            "\n",
            "--2020-08-22 07:21:52--  http://efrosgans.eecs.berkeley.edu/cyclegan/pretrained_models/style_cezanne.pth\n",
            "Resolving efrosgans.eecs.berkeley.edu (efrosgans.eecs.berkeley.edu)... 128.32.189.73\n",
            "Connecting to efrosgans.eecs.berkeley.edu (efrosgans.eecs.berkeley.edu)|128.32.189.73|:80... connected.\n",
            "HTTP request sent, awaiting response... 200 OK\n",
            "Length: 45575747 (43M)\n",
            "Saving to: ‘./checkpoints/style_cezanne_pretrained/latest_net_G.pth’\n",
            "\n",
            "./checkpoints/style 100%[===================>]  43.46M  13.7MB/s    in 3.3s    \n",
            "\n",
            "2020-08-22 07:21:56 (13.1 MB/s) - ‘./checkpoints/style_cezanne_pretrained/latest_net_G.pth’ saved [45575747/45575747]\n",
            "\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "lLQI4pBnDZxu",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 107
        },
        "outputId": "77ddc6c6-14e9-4e04-df0b-5ee29f5214bb"
      },
      "source": [
        "! pip install dominate"
      ],
      "execution_count": 158,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Collecting dominate\n",
            "  Downloading https://files.pythonhosted.org/packages/4f/e6/794a119963b7cfe4bd41177c8f9d4195fe901652f04189fbd2edf513c7b2/dominate-2.5.1-py2.py3-none-any.whl\n",
            "Installing collected packages: dominate\n",
            "Successfully installed dominate-2.5.1\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "W-jRJ39EFi2J",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 35
        },
        "outputId": "c0276ed7-ef31-473a-e973-9ae8e3da2222"
      },
      "source": [
        "% cd /content/\n",
        "! mkdir testimages"
      ],
      "execution_count": 168,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "/content\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "kbhB7cilFoqY",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 315
        },
        "outputId": "33f19fda-b4de-421f-ddbc-b528d62d3122"
      },
      "source": [
        "! wget https://img.bobowin.blog/uploads/20180627140424_92.jpg\n",
        "! mv 20180627140424_92.jpg testimages/"
      ],
      "execution_count": 169,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "--2020-08-22 07:34:50--  https://img.bobowin.blog/uploads/20180627140424_92.jpg\n",
            "Resolving img.bobowin.blog (img.bobowin.blog)... 172.105.210.73\n",
            "Connecting to img.bobowin.blog (img.bobowin.blog)|172.105.210.73|:443... connected.\n",
            "HTTP request sent, awaiting response... 301 Moved Permanently\n",
            "Location: https://i2.achangpro.com/img.bobowin.blog/uploads/20180627140424_92.jpg [following]\n",
            "--2020-08-22 07:34:52--  https://i2.achangpro.com/img.bobowin.blog/uploads/20180627140424_92.jpg\n",
            "Resolving i2.achangpro.com (i2.achangpro.com)... 172.67.5.120, 104.22.7.240, 104.22.6.240, ...\n",
            "Connecting to i2.achangpro.com (i2.achangpro.com)|172.67.5.120|:443... connected.\n",
            "HTTP request sent, awaiting response... 200 OK\n",
            "Length: 141445 (138K) [image/jpeg]\n",
            "Saving to: ‘20180627140424_92.jpg’\n",
            "\n",
            "20180627140424_92.j 100%[===================>] 138.13K   608KB/s    in 0.2s    \n",
            "\n",
            "2020-08-22 07:34:53 (608 KB/s) - ‘20180627140424_92.jpg’ saved [141445/141445]\n",
            "\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "8gIp9HJGAbFX",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 884
        },
        "outputId": "a340ce4e-e5b3-4bcf-8c4c-28a46b5f3c88"
      },
      "source": [
        "% cd /content/pytorch-CycleGAN-and-pix2pix\n",
        "! python test.py --dataroot /content/testimages --name style_cezanne_pretrained --model test --no_dropout\n"
      ],
      "execution_count": 170,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "/content/pytorch-CycleGAN-and-pix2pix\n",
            "----------------- Options ---------------\n",
            "             aspect_ratio: 1.0                           \n",
            "               batch_size: 1                             \n",
            "          checkpoints_dir: ./checkpoints                 \n",
            "                crop_size: 256                           \n",
            "                 dataroot: /content/testimages           \t[default: None]\n",
            "             dataset_mode: single                        \n",
            "                direction: AtoB                          \n",
            "          display_winsize: 256                           \n",
            "                    epoch: latest                        \n",
            "                     eval: False                         \n",
            "                  gpu_ids: 0                             \n",
            "                init_gain: 0.02                          \n",
            "                init_type: normal                        \n",
            "                 input_nc: 3                             \n",
            "                  isTrain: False                         \t[default: None]\n",
            "                load_iter: 0                             \t[default: 0]\n",
            "                load_size: 256                           \n",
            "         max_dataset_size: inf                           \n",
            "                    model: test                          \n",
            "             model_suffix:                               \n",
            "               n_layers_D: 3                             \n",
            "                     name: style_cezanne_pretrained      \t[default: experiment_name]\n",
            "                      ndf: 64                            \n",
            "                     netD: basic                         \n",
            "                     netG: resnet_9blocks                \n",
            "                      ngf: 64                            \n",
            "               no_dropout: True                          \t[default: False]\n",
            "                  no_flip: False                         \n",
            "                     norm: instance                      \n",
            "                 num_test: 50                            \n",
            "              num_threads: 4                             \n",
            "                output_nc: 3                             \n",
            "                    phase: test                          \n",
            "               preprocess: resize_and_crop               \n",
            "              results_dir: ./results/                    \n",
            "           serial_batches: False                         \n",
            "                   suffix:                               \n",
            "                  verbose: False                         \n",
            "----------------- End -------------------\n",
            "dataset [SingleDataset] was created\n",
            "initialize network with normal\n",
            "model [TestModel] was created\n",
            "loading the model from ./checkpoints/style_cezanne_pretrained/latest_net_G.pth\n",
            "---------- Networks initialized -------------\n",
            "[Network G] Total number of parameters : 11.378 M\n",
            "-----------------------------------------------\n",
            "creating web directory ./results/style_cezanne_pretrained/test_latest\n",
            "processing (0000)-th image... ['/content/testimages/20180627140424_92.jpg']\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "S0XjF0t9Dcrq",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 381
        },
        "outputId": "29c40949-0b53-40a6-dce4-1cf98e5e4e80"
      },
      "source": [
        "! ls ./results/style_cezanne_pretrained/test_latest/images"
      ],
      "execution_count": 173,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "10_fake.png  1_real.png\t\t\t 28_fake.png  37_real.png  47_fake.png\n",
            "10_real.png  2017-03-17_fake.png\t 28_real.png  38_fake.png  47_real.png\n",
            "11_fake.png  2017-03-17_real.png\t 29_fake.png  38_real.png  48_fake.png\n",
            "11_real.png  20180627140424_92_fake.png  29_real.png  39_fake.png  48_real.png\n",
            "12_fake.png  20180627140424_92_real.png  2_fake.png   39_real.png  49_fake.png\n",
            "12_real.png  20_fake.png\t\t 2_real.png   3_fake.png   49_real.png\n",
            "13_fake.png  20_real.png\t\t 30_fake.png  3_real.png   4_fake.png\n",
            "13_real.png  21_fake.png\t\t 30_real.png  40_fake.png  4_real.png\n",
            "14_fake.png  21_real.png\t\t 31_fake.png  40_real.png  50_fake.png\n",
            "14_real.png  22_fake.png\t\t 31_real.png  41_fake.png  50_real.png\n",
            "15_fake.png  22_real.png\t\t 32_fake.png  41_real.png  51_fake.png\n",
            "15_real.png  23_fake.png\t\t 32_real.png  42_fake.png  51_real.png\n",
            "16_fake.png  23_real.png\t\t 33_fake.png  42_real.png  52_fake.png\n",
            "16_real.png  24_fake.png\t\t 33_real.png  43_fake.png  52_real.png\n",
            "17_fake.png  24_real.png\t\t 34_fake.png  43_real.png  53_fake.png\n",
            "17_real.png  25_fake.png\t\t 34_real.png  44_fake.png  53_real.png\n",
            "18_fake.png  25_real.png\t\t 35_fake.png  44_real.png  5_fake.png\n",
            "18_real.png  26_fake.png\t\t 35_real.png  45_fake.png  5_real.png\n",
            "19_fake.png  26_real.png\t\t 36_fake.png  45_real.png\n",
            "19_real.png  27_fake.png\t\t 36_real.png  46_fake.png\n",
            "1_fake.png   27_real.png\t\t 37_fake.png  46_real.png\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "GxZGbOoBGIDF",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 273
        },
        "outputId": "87e94c18-3f1a-4028-a4a2-e36659b2aac5"
      },
      "source": [
        "from PIL import Image\n",
        "Image.open('./results/style_cezanne_pretrained/test_latest/images/20180627140424_92_fake.png')"
      ],
      "execution_count": 174,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<PIL.PngImagePlugin.PngImageFile image mode=RGB size=256x256 at 0x7F6E0409AA90>"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 174
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "46CK-NOtGORv",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 273
        },
        "outputId": "3f07f9a0-2d8b-4205-ef78-b4bdc8f8f88d"
      },
      "source": [
        "from PIL import Image\n",
        "Image.open('./results/style_cezanne_pretrained/test_latest/images/20180627140424_92_real.png')"
      ],
      "execution_count": 175,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<PIL.PngImagePlugin.PngImageFile image mode=RGB size=256x256 at 0x7F6E04062A20>"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 175
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "ySwVYTEPV2EE",
        "colab_type": "text"
      },
      "source": [
        "## 強化學習框架\n",
        "- https://github.com/ywchiu/pydeeplearning/blob/master/chapter7/7.11%20Using%20Keras%20and%20Deep%20Q-Network%20to%20Play%20FlappyBird.ipynb"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "CbqVgOD2alTR",
        "colab_type": "text"
      },
      "source": [
        "## ROC Curve"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "hXZkfSkpGSv-",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "from sklearn.datasets import load_iris\n",
        "iris = load_iris()\n",
        "X = iris.data[50:,]\n",
        "y = iris.target[50:]"
      ],
      "execution_count": 1,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "5tpj3fW1a8YA",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "from sklearn.preprocessing import LabelEncoder\n",
        "le = LabelEncoder()\n",
        "y = le.fit_transform(y)"
      ],
      "execution_count": 3,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "aaMzImmwbH7Y",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 104
        },
        "outputId": "c580b569-141f-4dd1-a70b-e1ee6ba83f2e"
      },
      "source": [
        "y"
      ],
      "execution_count": 4,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
              "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
              "       0, 0, 0, 0, 0, 0, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\n",
              "       1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\n",
              "       1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1])"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 4
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "hAgM1xWRa14f",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 104
        },
        "outputId": "0933291a-25da-44e8-c559-6dbdf6832055"
      },
      "source": [
        "from sklearn.linear_model import LogisticRegression\n",
        "clf = LogisticRegression()\n",
        "clf.fit(X,y)"
      ],
      "execution_count": 5,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "LogisticRegression(C=1.0, class_weight=None, dual=False, fit_intercept=True,\n",
              "                   intercept_scaling=1, l1_ratio=None, max_iter=100,\n",
              "                   multi_class='auto', n_jobs=None, penalty='l2',\n",
              "                   random_state=None, solver='lbfgs', tol=0.0001, verbose=0,\n",
              "                   warm_start=False)"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 5
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "m7S12HOibeC9",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1000
        },
        "outputId": "2a1a8303-96a2-4995-b24f-43e00515ad97"
      },
      "source": [
        "pred_y"
      ],
      "execution_count": 9,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "array([[8.42360997e-01, 1.57639003e-01],\n",
              "       [8.56151658e-01, 1.43848342e-01],\n",
              "       [6.80762131e-01, 3.19237869e-01],\n",
              "       [9.48609329e-01, 5.13906708e-02],\n",
              "       [7.89970560e-01, 2.10029440e-01],\n",
              "       [8.56397624e-01, 1.43602376e-01],\n",
              "       [7.24664302e-01, 2.75335698e-01],\n",
              "       [9.95957523e-01, 4.04247739e-03],\n",
              "       [8.69770011e-01, 1.30229989e-01],\n",
              "       [9.54950163e-01, 4.50498367e-02],\n",
              "       [9.91463507e-01, 8.53649327e-03],\n",
              "       [9.13965415e-01, 8.60345849e-02],\n",
              "       [9.77828834e-01, 2.21711656e-02],\n",
              "       [7.62589964e-01, 2.37410036e-01],\n",
              "       [9.88286676e-01, 1.17133236e-02],\n",
              "       [9.15712566e-01, 8.42874344e-02],\n",
              "       [7.96650328e-01, 2.03349672e-01],\n",
              "       [9.75185894e-01, 2.48141056e-02],\n",
              "       [7.67014851e-01, 2.32985149e-01],\n",
              "       [9.78837218e-01, 2.11627820e-02],\n",
              "       [4.95533661e-01, 5.04466339e-01],\n",
              "       [9.67979492e-01, 3.20205078e-02],\n",
              "       [5.52964425e-01, 4.47035575e-01],\n",
              "       [8.31880444e-01, 1.68119556e-01],\n",
              "       [9.36983297e-01, 6.30167028e-02],\n",
              "       [9.08437107e-01, 9.15628931e-02],\n",
              "       [7.50003624e-01, 2.49996376e-01],\n",
              "       [4.62714210e-01, 5.37285790e-01],\n",
              "       [8.13351220e-01, 1.86648780e-01],\n",
              "       [9.95221773e-01, 4.77822747e-03],\n",
              "       [9.82646004e-01, 1.73539956e-02],\n",
              "       [9.89760453e-01, 1.02395473e-02],\n",
              "       [9.77554355e-01, 2.24456453e-02],\n",
              "       [3.47312323e-01, 6.52687677e-01],\n",
              "       [7.83571581e-01, 2.16428419e-01],\n",
              "       [8.15607281e-01, 1.84392719e-01],\n",
              "       [7.79806700e-01, 2.20193300e-01],\n",
              "       [8.86832320e-01, 1.13167680e-01],\n",
              "       [9.53522631e-01, 4.64773687e-02],\n",
              "       [9.53389356e-01, 4.66106444e-02],\n",
              "       [8.94629375e-01, 1.05370625e-01],\n",
              "       [8.19258792e-01, 1.80741208e-01],\n",
              "       [9.68615549e-01, 3.13844509e-02],\n",
              "       [9.95909393e-01, 4.09060664e-03],\n",
              "       [9.29178718e-01, 7.08212823e-02],\n",
              "       [9.52991166e-01, 4.70088344e-02],\n",
              "       [9.37978081e-01, 6.20219188e-02],\n",
              "       [9.32161700e-01, 6.78382998e-02],\n",
              "       [9.98119961e-01, 1.88003893e-03],\n",
              "       [9.50639610e-01, 4.93603902e-02],\n",
              "       [6.57687427e-03, 9.93423126e-01],\n",
              "       [1.92340464e-01, 8.07659536e-01],\n",
              "       [2.66970920e-02, 9.73302908e-01],\n",
              "       [8.63787264e-02, 9.13621274e-01],\n",
              "       [2.22822897e-02, 9.77717710e-01],\n",
              "       [4.27600765e-03, 9.95723992e-01],\n",
              "       [5.86458271e-01, 4.13541729e-01],\n",
              "       [1.77097998e-02, 9.82290200e-01],\n",
              "       [4.77722033e-02, 9.52227797e-01],\n",
              "       [8.14879672e-03, 9.91851203e-01],\n",
              "       [2.41615883e-01, 7.58384117e-01],\n",
              "       [1.43765004e-01, 8.56234996e-01],\n",
              "       [7.29542800e-02, 9.27045720e-01],\n",
              "       [1.78638368e-01, 8.21361632e-01],\n",
              "       [6.96506470e-02, 9.30349353e-01],\n",
              "       [7.62413399e-02, 9.23758660e-01],\n",
              "       [1.26155589e-01, 8.73844411e-01],\n",
              "       [3.93003034e-03, 9.96069970e-01],\n",
              "       [9.30414923e-04, 9.99069585e-01],\n",
              "       [4.12720552e-01, 5.87279448e-01],\n",
              "       [3.01887683e-02, 9.69811232e-01],\n",
              "       [2.46384858e-01, 7.53615142e-01],\n",
              "       [3.81484940e-03, 9.96185151e-01],\n",
              "       [3.98959251e-01, 6.01040749e-01],\n",
              "       [4.68071137e-02, 9.53192886e-01],\n",
              "       [4.64415775e-02, 9.53558423e-01],\n",
              "       [4.73812854e-01, 5.26187146e-01],\n",
              "       [4.17095379e-01, 5.82904621e-01],\n",
              "       [4.32861516e-02, 9.56713848e-01],\n",
              "       [1.13537980e-01, 8.86462020e-01],\n",
              "       [2.45255923e-02, 9.75474408e-01],\n",
              "       [1.64046412e-02, 9.83595359e-01],\n",
              "       [3.43109975e-02, 9.65689002e-01],\n",
              "       [4.45340928e-01, 5.54659072e-01],\n",
              "       [1.64554303e-01, 8.35445697e-01],\n",
              "       [1.17853741e-02, 9.88214626e-01],\n",
              "       [2.78611457e-02, 9.72138854e-01],\n",
              "       [1.27471504e-01, 8.72528496e-01],\n",
              "       [4.79741910e-01, 5.20258090e-01],\n",
              "       [1.03555866e-01, 8.96444134e-01],\n",
              "       [2.79640214e-02, 9.72035979e-01],\n",
              "       [1.46476475e-01, 8.53523525e-01],\n",
              "       [1.92340464e-01, 8.07659536e-01],\n",
              "       [1.63792563e-02, 9.83620744e-01],\n",
              "       [1.83321006e-02, 9.81667899e-01],\n",
              "       [1.01033229e-01, 8.98966771e-01],\n",
              "       [2.59756050e-01, 7.40243950e-01],\n",
              "       [1.76597808e-01, 8.23402192e-01],\n",
              "       [5.93089647e-02, 9.40691035e-01],\n",
              "       [2.68991559e-01, 7.31008441e-01]])"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 9
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "hHujP4nEbNZW",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "from sklearn.metrics import roc_curve\n",
        "pred_y = clf.predict_proba(X)\n",
        "fpr, tpr, threshold = roc_curve(y, pred_y[:,1])"
      ],
      "execution_count": 10,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Uu2u6kbzbUPW",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 283
        },
        "outputId": "6cd6d505-c3af-4262-a17d-823d0be73ad6"
      },
      "source": [
        "from matplotlib import pyplot as plt\n",
        "plt.plot(fpr, tpr)"
      ],
      "execution_count": 11,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "[<matplotlib.lines.Line2D at 0x7fb4570a33c8>]"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 11
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "dJZe8Qf4bpJV",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 283
        },
        "outputId": "d61b17ac-eb6c-432c-8fb3-48eebf59afd1"
      },
      "source": [
        "plt.plot(threshold, 1 - fpr)\n",
        "plt.plot(threshold, tpr)"
      ],
      "execution_count": 14,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "[<matplotlib.lines.Line2D at 0x7fb4567a7400>]"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 14
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "dmnaia6sb4QT",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        ""
      ],
      "execution_count": null,
      "outputs": []
    }
  ]
}